SlideShare a Scribd company logo
1 of 58
Download to read offline
Emulators for models and complexity reduction
Akil Narayan1
1
Department of Mathematics, and Scientific Computing and Imaging (SCI) Institute
University of Utah
August 2018
SAMSI MUMS opening workshop
A. Narayan (U. Utah) Emulators and surrogates
Models and emulators
y = u(x) + ε
x ∈ D ⊆ Rd
y ∈ RP
x
y
The parameters/factors x govern the bulk behavior of the response u
The noise or error ε can account for model discrepancy
The observable y can be deterministic or stochastic
A. Narayan (U. Utah) Emulators and surrogates
Models and emulators
y = u(x) + ε
x ∈ D ⊆ Rd
y ∈ RP
x
y
Available data: noisy measurements, y, abstractly treated as samples at
specific values of x.
Emulators are generally built to be consistent with data. Their purpose can
be to
extrapolate/interpolate data
accelerate queries of the model
analyze for variances, screening, sensitivity, etc.
A. Narayan (U. Utah) Emulators and surrogates
Models and emulators
y = u(x) + ε
x ∈ D ⊆ Rd
y ∈ RP
x
y
I will primarily discuss emulator constructions from applied
mathematics/scientific computing.
We are interested in things like stability, accuracy, consistency, etc.
Take-home point: experimental design is critical in building good emulators.
A. Narayan (U. Utah) Emulators and surrogates
Building emulators
Many mathematical emulator models have the form
u(x) ≈ uN (x) :=
N
n=1
cnφn(x),
Information about y: sample data (xm, ym), m = 1, . . . , M.
Two general types of approximations:
linear approximations: uN is linear in the data.
φn(·) are prescribed a priori, {ym} → {cn} is linear
nonlinear approximations: uN is nonlinearly dependent on data
computation of cn may be nonlinear
identification of φn may depend on data
A. Narayan (U. Utah) Emulators and surrogates
Building emulators
Many mathematical emulator models have the form
u(x) ≈ uN (x) :=
N
n=1
cnφn(x),
Information about y: sample data (xm, ym), m = 1, . . . , M.
Two general types of approximations:
linear approximations: uN is linear in the data.
φn(·) are prescribed a priori, {ym} → {cn} is linear
nonlinear approximations: uN is nonlinearly dependent on data
computation of cn may be nonlinear
identification of φn may depend on data
The form of φn does not generally dictate linear/nonlinearity.
Some linear approximations:
interpolation
quadrature
least-squares
Some nonlinear approximations:
radial basis/kernel approximations
non-quadratic regularized approximation
proper orthogonal decomposition
A. Narayan (U. Utah) Emulators and surrogates
Building emulators
Many mathematical emulator models have the form
u(x) ≈ uN (x) :=
N
n=1
cnφn(x),
Information about y: sample data (xm, ym), m = 1, . . . , M.
Two general types of approximations:
linear approximations: uN is linear in the data.
φn(·) are prescribed a priori, {ym} → {cn} is linear
nonlinear approximations: uN is nonlinearly dependent on data
computation of cn may be nonlinear
identification of φn may depend on data
Example: If M ≥ N, cj coefficients computable via least-squares





y1
y2
...
yM





= y ≈ Ac =





φ1(x1) φ2(x1) · · · φN (x1)
φ1(x2) φ2(x2) · · · φN (x2)
...
...
...
...
φ1(xM ) φ2(xM ) · · · φN (xM )










c1
c2
...
cN





A. Narayan (U. Utah) Emulators and surrogates
Emulators as model reduction
Emulators are built in the hope that x → u(x) is a map of low complexity.
If true, and an efficient model to capture this complexity is discoverable, then
u(x) ≈ uN (x) =
N
n=1
cnφn(x), V := span{φ1, . . . , φN }
can be achieved with “small” N.
A. Narayan (U. Utah) Emulators and surrogates
Emulators as model reduction
Emulators are built in the hope that x → u(x) is a map of low complexity.
If true, and an efficient model to capture this complexity is discoverable, then
u(x) ≈ uN (x) =
N
n=1
cnφn(x), V := span{φ1, . . . , φN }
can be achieved with “small” N.
Identify V
Efficiently construct uN from V
Neither of these is particularly easy in general.
Anyway, scientific models are complex, is this even feasible with reasonable
N?
A. Narayan (U. Utah) Emulators and surrogates
An explicit example
Example: Consider the solution u(z; x) to the parameterized PDE:
− z · (a(z; x) zu(z; x)) = f(z), (z, x) ∈ Ω × D,
u(z; x) = 0, (z, x) ∈ ∂Ω × D.
For each x, u(·; x) ∈ H = H1
(Ω). Let the diffusion coefficient be given by
a(z; x) =
∞
j=1
xjψj(z).
A. Narayan (U. Utah) Emulators and surrogates
An explicit example
Example: Consider the solution u(z; x) to the parameterized PDE:
− z · (a(z; x) zu(z; x)) = f(z), (z, x) ∈ Ω × D,
u(z; x) = 0, (z, x) ∈ ∂Ω × D.
For each x, u(·; x) ∈ H = H1
(Ω). Let the diffusion coefficient be given by
a(z; x) =
∞
j=1
xjψj(z).
If x = (x1, . . .) ∈ D = [−1, 1]∞
, and there is some p ≤ 1 such that
∞
j=1
ψj
p
L∞(Ω) < ∞,
then an emulator uN can be constructed such that
u − uN L2(D,H) N−r
, r =
1
p
−
1
2
.
[Cohen, DeVore, Schwab 2010]
A. Narayan (U. Utah) Emulators and surrogates
Adapted vs linear
An approximation to u:
u ≈ uN (z; x) =
N
n=1
cn(z)φn(x), V := span{φ1, . . . , φN }
Non-adapted approximation: With V chosen, construct uN so that
u − uN L2(D,RP ) inf
v∈V
u − v L2(D,RP )
The main task is to compute uN from a given V .
A. Narayan (U. Utah) Emulators and surrogates
Adapted vs linear
An approximation to u:
u ≈ uN (z; x) =
N
n=1
cn(z)φn(x), V := span{φ1, . . . , φN }
Non-adapted approximation: With V chosen, construct uN so that
u − uN L2(D,RP ) inf
v∈V
u − v L2(D,RP )
The main task is to compute uN from a given V .
Adapated approximation: Find V and uN so that
u(x) − uN (x) RP is “small” for all x ∈ D
A. Narayan (U. Utah) Emulators and surrogates
Adapted vs linear
An approximation to u:
u ≈ uN (z; x) =
N
n=1
cn(z)φn(x), V := span{φ1, . . . , φN }
Non-adapted approximation: With V chosen, construct uN so that
u − uN L2(D,RP ) inf
v∈V
u − v L2(D,RP )
The main task is to compute uN from a given V .
Adapated approximation: Find V and uN so that
u(x) − uN (x) RP is “small” for all x ∈ D
Adapted approximations are always nonlinear.
Non-adapted approximations can be linear.
A. Narayan (U. Utah) Emulators and surrogates
Emulators and sampling/experimental design
y = u +
u ≈ uN =
N
n=1
cnφn(x) ∈ V,
{(xm, ym)}
M
m=1 −→ {cn}
N
n=1
Desiderata:
u − uN B small for a normed vector space B
M of “reasonable” size
Accuracy, both in identification of V and in computation of uN depends
largely on sample design, i.e., the choice of x1, . . . , xM .
A. Narayan (U. Utah) Emulators and surrogates
Emulators and sampling/experimental design
y = u +
u ≈ uN =
N
n=1
cnφn(x) ∈ V,
{(xm, ym)}
M
m=1 −→ {cn}
N
n=1
Desiderata:
u − uN B small for a normed vector space B
M of “reasonable” size
Accuracy, both in identification of V and in computation of uN depends
largely on sample design, i.e., the choice of x1, . . . , xM .
Good sample design can minimize required data size M
Intelligent sampling enables efficient emulator construction
A. Narayan (U. Utah) Emulators and surrogates
Summary of methods
We’ll see how sampling design affects approximation statements for three
strategies:
Discrete least-squares: linear approximation, M ≥ N
Compressive sampling: nonlinear approximation, M N
Reduced order modeling: nonlinear approximation, N ∼ M = O(1)
I’ll discuss optimal mathematical statements one can make, taking the form
u − uN B KN × (Best approx error) + if M ≥ KM .
I will focus on the role that sampling plays in these techniques.
A. Narayan (U. Utah) Emulators and surrogates
Summary of methods
We’ll see how sampling design affects approximation statements for three
strategies:
Discrete least-squares: linear approximation, M ≥ N
Compressive sampling: nonlinear approximation, M N
Reduced order modeling: nonlinear approximation, N ∼ M = O(1)
I’ll discuss optimal mathematical statements one can make, taking the form
u − uN B KN × (Best approx error) + if M ≥ KM .
I will focus on the role that sampling plays in these techniques.
Warning: There are entire sub-fields of applied math and statistics
concerning sampling that I will ignore.
A. Narayan (U. Utah) Emulators and surrogates
Summary of methods
We’ll see how sampling design affects approximation statements for three
strategies:
Discrete least-squares: linear approximation, M ≥ N
Compressive sampling: nonlinear approximation, M N
Reduced order modeling: nonlinear approximation, N ∼ M = O(1)
I’ll discuss optimal mathematical statements one can make, taking the form
u − uN B KN × (Best approx error) + if M ≥ KM .
I will focus on the role that sampling plays in these techniques.
Warning: There are entire sub-fields of applied math and statistics
concerning sampling that I will ignore.
(Because they’re not directly relevant to the message.)
A. Narayan (U. Utah) Emulators and surrogates
Part I: Linear approximation
Discrete least squares
Non-adapted basis functions, linear approximation construction procedure
A. Narayan (U. Utah) Emulators and surrogates
An aside – polynomials and PCE
u(x) + = y ≈ uN (x) =
N
n=1
cnφn(x),
x ∈ D ⊆ Rd
y ∈ R c ∈ RN
Non-adapted approximation: the φn are a priori chosen.
We often choose polynomials. (Cf. PCE)
Why?
A. Narayan (U. Utah) Emulators and surrogates
An aside – polynomials and PCE
u(x) + = y ≈ uN (x) =
N
n=1
cnφn(x),
x ∈ D ⊆ Rd
y ∈ R c ∈ RN
Non-adapted approximation: the φn are a priori chosen.
We often choose polynomials. (Cf. PCE)
Why?
Polynomials are easy to compute with/evaluate
Polynomial expansions are (reasonably) easy to manipulate, multiply,
differentiate, etc.
Polynomials provide best approximation numbers that behave optimally:
inf
N=dim P =dim P d
k
P ⊂Hs
sup
f∈Hs
f Hs =1
inf
p∈P
f − p Hs ∼ N−s/d
sup
f∈Hs
f Hs =1
inf
p∈P d
k
N=dim P d
k
f − p Hs N−s/d
[Pinkus 1985]
A. Narayan (U. Utah) Emulators and surrogates
Mathematical preliminaries
u(x) + = y ≈ uN (x) =
N
n=1
cnφn(x),
x ∈ D ⊆ Rd
y ∈ R c ∈ RN
Least-squares problem is approximation of the form
(c∗
1, . . . , c∗
N )
T
= c∗
= arg min
c∈RN
M
m=1
[uN (xm) − ym]
2
A. Narayan (U. Utah) Emulators and surrogates
Mathematical preliminaries
u(x) + = y ≈ uN (x) =
N
n=1
cnφn(x),
x ∈ D ⊆ Rd
y ∈ R c ∈ RN
Least-squares problem is approximation of the form
(c∗
1, . . . , c∗
N )
T
= c∗
= arg min
c∈RN
M
m=1
[uN (xm) − ym]
2
Let V := span {φ1, . . . , φN }. Least-squares is, equivalently,
v∗
= arg min
v∈V
M
m=1
(v(xm) − ym)
2
V is an a priori space of functions.
What is the “best” approximation we can hope for?
A. Narayan (U. Utah) Emulators and surrogates
Mathematical preliminaries
Given a probability measure µ on D, approximation will take place in an L2
space:
g, h µ :=
D
g(x)h(x)dµ(x), L2
µ (D) := g : D → R g µ < ∞
The best approximation error to u from the subspace V is
σV (u) := inf
v∈V
u − v µ
A. Narayan (U. Utah) Emulators and surrogates
Mathematical preliminaries
Given a probability measure µ on D, approximation will take place in an L2
space:
g, h µ :=
D
g(x)h(x)dµ(x), L2
µ (D) := g : D → R g µ < ∞
The best approximation error to u from the subspace V is
σV (u) := inf
v∈V
u − v µ
Randomized sampling: xm sampled iid from µ, and no noise, ym = u(xm),
uN = arg min
v∈V
M
m=1
(v(xm) − ym)
2
Law of large numbers: M ↑ ∞ ⇒ uN − u µ → σV (u).
A. Narayan (U. Utah) Emulators and surrogates
“Standard” Monte Carlo
Approximate a function
u(x) = exp −ω x −
1
π
2
, x ∈ [−1, 1], = 0,
with µ uniform on [−1, 1], from the space of potential surrogates
V = span 1, . . . , xN−1
Data xm sampled iid from µ
Convergence observed, but slow
Why does this happen, and can we fix it?
50 100 150 200 250 300
10−5
10−3
10−1
101
103
105
107
M
Mean-squareerror
D = [−1, 1], N = 50
Optimal error
MC
A. Narayan (U. Utah) Emulators and surrogates
“Standard” Monte Carlo
Approximate a function
u(x) = exp −ω x −
1
π
2
, x ∈ [−1, 1], = 0,
with µ uniform on [−1, 1], from the space of potential surrogates
V = span 1, . . . , xN−1
Data xm sampled iid from µ
Convergence observed, but slow
Why does this happen, and can we fix it?
50 100 150 200 250 300
10−5
10−3
10−1
101
103
105
107
M
Mean-squareerror
D = [−1, 1], N = 50
Optimal error
MC
Sampling from a standard distribution is frequently suboptimal
A. Narayan (U. Utah) Emulators and surrogates
Convergence results
Proximity to the optimal solution is guaranteed with enough samples.
Define
Kµ(V ) := sup
x∈D
sup
v∈V {0}
|v(x)|2
v 2
µ
If x1, . . . , xM are sampled iid from µ, then
M
log M
≥
2 + 2r
log(e/2)
Kµ(V )
guarantees that, with probability ≥ 1 − 2M−r
,
E u − uN
2
µ ≤ 1 + 2
1 − log 2
(1 + r) log M
σV (u)2
+ 8U2
M−r
,
where U = supx∈D |u(x)|, and
uN = arg min
v∈V
M
m=1
(v(xm) − ym)
2
[Cohen, Davenport, Leviatan 2013]
A. Narayan (U. Utah) Emulators and surrogates
Randomized sampling – Monte Carlo
M
log M
≥
2 + 2r
log(e/2)
Kµ(V ), Kµ(V ) = sup
x∈D
sup
v∈V {0}
|v(x)|2
v 2
µ
The smallest (best) value of Kµ(V ) is N.
A. Narayan (U. Utah) Emulators and surrogates
Randomized sampling – Monte Carlo
M
log M
≥
2 + 2r
log(e/2)
Kµ(V ), Kµ(V ) = sup
x∈D
sup
v∈V {0}
|v(x)|2
v 2
µ
The smallest (best) value of Kµ(V ) is N.
Example: Linear models, N = d + 1
φ1(x) = 1, φj+1(x) = xj, j = 1, . . . , d
Let µ be the standard Gaussian measure over D = Rd
Then Kµ(V ) = ∞.
Analysis suggests this is a pretty bad sampling design, but in practice it’s fine.
A. Narayan (U. Utah) Emulators and surrogates
Randomized sampling – Monte Carlo
M
log M
≥
2 + 2r
log(e/2)
Kµ(V ), Kµ(V ) = sup
x∈D
sup
v∈V {0}
|v(x)|2
v 2
µ
The smallest (best) value of Kµ(V ) is N.
Example: Linear models, N = d + 1
φ1(x) = 1, φj+1(x) = xj, j = 1, . . . , d
Let µ be the standard Gaussian measure over D = Rd
Then Kµ(V ) = ∞.
Analysis suggests this is a pretty bad sampling design, but in practice it’s fine.
In the previous example, Kµ(V ) ∼ N2
.
In practice, Kµ(V ) depends exponentially on d.
The ideal case: Kµ(V ) ∼ N. To accomplish this, use biased sampling.
A. Narayan (U. Utah) Emulators and surrogates
Randomized sampling – weighted methods
Lesson: sampling xm ∼ µ is usually not optimal, and sometimes terrible.
Standard least-squares:
arg min
c
Ac − y 2
Weighted least-squares:
arg min
c
Ac − y 2,w = arg min
c
√
W Ac −
√
W y
2
where W = diag(w1, . . . , wM ) contains positive weights wj.
A. Narayan (U. Utah) Emulators and surrogates
Randomized sampling – optimality
We can entirely circumvent the Kµ(V ) problem by changing sampling
measures.
Assume φ1, . . . φN is an L2
µ-orthonormal basis for V . Generate x1, . . . , xM
iid from µV , where
dµV (x) =
1
N
N
n=1
φ2
n(x)dµ(x).
Use weights
wm =
dµ
dµV
(xm) =
N
N
n=1 φ2
n(xm)
.
Our weighted least-squares estimator is defined by
c∗
= arg min
c
Ac − y 2,w
The measure µV is called the induced distribution for V .
A. Narayan (U. Utah) Emulators and surrogates
Randomized sampling – optimality
Let x1, . . . , xM ∼ µV , with uN (x) =
N
n=1 c∗
nφn(x) computed via
c∗
= arg min
c
Ac − y 2,w
Then
M
log M
≥
2 + 2r
log(e/2)
N
guarantees that, with probability ≥ 1 − 2M−r
,
E u − uN
2
µ ≤ 1 + 2
1 − log 2
(1 + r) log M
σV (µ)2
+ 8U2
M−r
[Cohen, Migliorati 2017]
Note: This M/N dependence is essentially optimal.
A. Narayan (U. Utah) Emulators and surrogates
The induced distribution
The induced distribution µV can be substantially different from µ.
x ∈ D = R2
, dµ(x) ∝ exp(− x 2
2),
x = (x(1)
, x(2)
), V = span x(1)
α1
x(2)
α2
(α1 + 1)(α2 + 1) ≤ 26
−6 −5 −4 −3 −2 −1 0 1 2 3 4 5 6
−6
−5
−4
−3
−2
−1
0
1
2
3
4
5
6
x(1)
x(2)
Samples from µ
−6 −5 −4 −3 −2 −1 0 1 2 3 4 5 6
−6
−5
−4
−3
−2
−1
0
1
2
3
4
5
6
x(1)
x(2)
Samples from µV
Under certain conditions, can sample from this distribution very efficiently, in
particular with linear complexity in d. [AN 2017]
A. Narayan (U. Utah) Emulators and surrogates
Randomized sampling – examples
This analysis tends to give accurate estimates
10 20 30 40 50
10−1
100
101
M
L2
µerror
N = 5
Optimal error = ε µ
µ
µV
50 100 150 200 250 300
10−5
10−3
10−1
101
103
105
107
M
L2
µerror
N = 45
A. Narayan (U. Utah) Emulators and surrogates
Randomized sampling – examples
This analysis tends to give accurate estimates
10 20 30 40 50
10−1
100
101
M
L2
µerror
N = 5
Optimal error = ε µ
µ
µV
50 100 150 200 250 300
10−5
10−3
10−1
101
103
105
107
M
L2
µerror
N = 45
Moral of the story:
randomized sampling according to µ is generally bad
randomized sampling according to µV is generally good
Intelligent sampling allows efficient, near-optimal computation of emulators.
A. Narayan (U. Utah) Emulators and surrogates
Odds and ends
Robust and accurate least-squares emulators for linear approximations can be
built with biased sampling.
Estimates are optimal: M N implies u − uN µ σV (u).
Estimates are d-independent.
Sampling is efficient if both µ and φn are tensor-product.
Convergence results robust to noise > 0.
No significant changes if y is vector-/function-valued
A. Narayan (U. Utah) Emulators and surrogates
Part II: Nonlinear approximation
Sparse approximation
Non-adapted basis functions, nonlinear approximation construction procedure
A. Narayan (U. Utah) Emulators and surrogates
Limited measurements
u(x) + = y ≈ uN (x) =
N
n=1
cnφn(x),
x ∈ D ⊆ Rd
y ∈ R c ∈ RN
V = span {φ1, . . . , φN }
{(xm, ym)}
M
m=1 → {cn}
N
n=1
When d > 1, it is common for an a priori N = dim V to be very large.
Least-squares: collecting M ∼ dim V measurements can be infeasible.
A. Narayan (U. Utah) Emulators and surrogates
Limited measurements
u(x) + = y ≈ uN (x) =
N
n=1
cnφn(x),
x ∈ D ⊆ Rd
y ∈ R c ∈ RN
V = span {φ1, . . . , φN }
{(xm, ym)}
M
m=1 → {cn}
N
n=1
When d > 1, it is common for an a priori N = dim V to be very large.
Least-squares: collecting M ∼ dim V measurements can be infeasible.
What happens when M < N? The system
Ac ≈ y,
is now underdetermined. Unique solutions can be gauranteed if functional
structure is imposed.
A. Narayan (U. Utah) Emulators and surrogates
Sparse approximation
uN (x) =
N
n=1
cnφn(x)
If M < N measurements are available, can we recover the largest M
coefficients from the vector c?
Assume
y(x) =
N
n=1
cnφn(x) + (x), | | < η.
The compressibility of y is measured by
σV,s(c) = inf
d∈RN
d 0≤s
c − d 1 , d 0 := {j ∈ {1, . . . , N} dj = 0
A. Narayan (U. Utah) Emulators and surrogates
Sparse approximation
y(x) =
N
n=1
cnφn(x) + (x), | | < η.
ym = y(xm), A ≈ cy
y is assumed to be compressible (i.e., c is assumed compressible)
With a limited number, M, of measurements, seek to approximate the
best s-term approximation of c.
Ideally, s ∼ M.
This is not possible if the sampling points are arbitrarily chosen.
A. Narayan (U. Utah) Emulators and surrogates
Compressed sensing
It is possible to recover the best s-term approximation with high probability.
Assume xm are iid sampled from µ, and that φn are L2
µ-orthonormal, and
assume
M CKµs log3
(s) log N,
For any c ∈ RN
, let ym = y(xm) =
N
n=1 cnφn(xm) + (xm), and assume
| | ≤ η.
Then, with probability exceeding 1 − N−γ log3
(s)
, the solution c∗
to the
optimization problem,
min d 1 such that Ad − y 2 ≤ η
√
M,
satisfies
c − c∗
1 ≤ C1σV,s(c) + C2
√
s .
Above,
Kµ = max
n=1,...,N
φn L∞(D).
[Rauhut 2010], [Rauhut, Ward 2010]
A. Narayan (U. Utah) Emulators and surrogates
Recovery of models with sparse representations
Figure 2. Transition plots for uniform random variables for d = 2 (top row
row). The left column corresponds to sampling from the random variable
column the CSA method and the right column asymptotic sampling.
For all low-dimensional and high-degree situations considered, CSA
and performs significantly better than than probabilistic sampling acco
optimization tolerances, and when th
tolerance, the authors of [23] obtained
We observe poor recovery since Kµ in sample requirement is poorly behaved:
M CKµs log3
(s) log N, Kµ = max
n=1,...,N
φn L∞(D).
This requirement is heavily dependent on µ.
A. Narayan (U. Utah) Emulators and surrogates
Better sampling
Again, choosing a better sampling strategy ameliorates this issue.
Sample xm ∼ µV , solve
min d 1 such that Ad − y 2,w ≤ η
√
M,
where w are weights to make the discrete sampling unbiased.
Exponential variables, Moreover the error in the approximation recovered by the asymptotic bounded
sampling method for Beta variables increases with dimension. When d = 30 the asymptotic bounded
sampling method fails to recover any polynomials regardless of the sparsity or the number of samples
used.
It is worth noting that case of Legendre polynomials sampled by Chebyshev distribution we have
a complete independence of the order of approximation, which agrees with previous results in [42].
However there are numerical results in [23, 49] showing almost no recovery when using the Chebyshev
sampling method in high-dimensions.
With the help of the authors of [23] we have verified that the poor performance exhibited in
the aforementioned papers is a result of numerical issues associated with the authors use of the
`1
-minimization solver in SparseLab [16]. Specifically, the authors of [23] were using more lenient
optimization tolerances, and when these tolerances were made tighter to match our optimization
tolerance, the authors of [23] obtained results consistent with Figure 2.
sampling method for Beta variables increases with dimension. When d = 30 the asymptotic bounded
sampling method fails to recover any polynomials regardless of the sparsity or the number of samples
used.
It is worth noting that case of Legendre polynomials sampled by Chebyshev distribution we have
a complete independence of the order of approximation, which agrees with previous results in [42].
However there are numerical results in [23, 49] showing almost no recovery when using the Chebyshev
sampling method in high-dimensions.
With the help of the authors of [23] we have verified that the poor performance exhibited in
the aforementioned papers is a result of numerical issues associated with the authors use of the
`1
-minimization solver in SparseLab [16]. Specifically, the authors of [23] were using more lenient
optimization tolerances, and when these tolerances were made tighter to match our optimization
tolerance, the authors of [23] obtained results consistent with Figure 2.
A GENERALIZED SAMPLING AND PRECONDITIONING SCHEME F
101
102
103
10 14
10 13
10 12
10 11
10 10
10 9
10 8
10 7
10 6
10 5
10 4
10 3
10 2
10 1
Number of samples M
`2error
CSA
MC
101
Figure 8. The e↵ect of dimension on the convergence of the CSA
the di↵usion equation (28). (Left) 30th degree polynomial in 2
polynomial in 20 dimensions
10 4
10 3
10 2
10 1
rror
CSA
Asymptotic
[Jakeman, AN 2017], [Guo, Zhou, Chen, AN 2016]
A. Narayan (U. Utah) Emulators and surrogates
Part III: Nonlinear approximation
Dimension reduction/reduced modeling
Adapted basis functions, nonlinear approximation construction procedure
A. Narayan (U. Utah) Emulators and surrogates
Dimension reduction
u(x) + = y ≈ uN (x) =
N
n=1
cnφn(x),
x ∈ D ⊆ Rd
y ∈ RP
cn ∈ RP
A “sample” ym is a vector, possible of large size, P 1.
In scientific models, P is also an indicator of the effort to obtain ym.
Construct V and φ1, . . . , φN , by analyzing
{(xm, ym)}
M
m=1 , (xm, ym) ∈ Rd
× RP
The φn are adapted to the data.
Though φn has no explicit form, evaluating such functions can be much
cheaper than gathering more data.
A. Narayan (U. Utah) Emulators and surrogates
Reduced basis methods
Gather (xm, ym) from a scientific model.
The reduced basis method (RBM) for nonlinear, adapted approximation,
constructs the emulator
uN (x) =
N
n=1
cnφn(x) =
N
n=1
yn n(x),
Here:
We need at least N = M data samples ym.
n are cardinal Lagrange functions, satisfying n(xm) = δn,m. They
have no explicit form.
The n are defined implicitly from the scientific model. (Via a Galerkin
procedure.)
This is not POD.
The space V = span{φn}N
n=1 is constructed/defined from the data and the
model.
There is no reason to believe this is a good idea unless xm is chosen well!
A. Narayan (U. Utah) Emulators and surrogates
Reduced basis methods
End goal: evaluation of surrogate uN should cost less than acquiring more
data. Costs:
Evaluting Lagrange functions n is the hard part – complexity usually
scales like N3
.
The full model ym is queried only at xm, and nowhere else.
Details of computational efficiency of the surrogate uN depend on
particular problem.
In practice, N ∼ O(10).
A. Narayan (U. Utah) Emulators and surrogates
Lagrange functions
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
µ
-3
-2
-1
0
1
2
The Lagrange functions
uN
10,1 uN
10,2 uN
10,3 uN
10,4 uN
10,5 uN
10,6 uN
10,7 uN
10,8 uN
10,9 uN
10,10
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
µ
2
4
6
8
˜∆10(µ)
5 10 15 20 25 30 35 40
Number of bases, N
10
-10
10
-5
10
0
-1 -0.5 0 0.5 1
µ
10
-6
10
-5
10
-4
10
-3
10
-2
10−4
× ˜∆10(µ)
||uN
(µ) − uN
10,E3
(µ)||X
S10
E3
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
µ
-1
-0.5
0
0.5
1
1.5
The Lagrange functions
uN
10,1 uN
10,2 uN
10,3 uN
10,4 uN
10,5 uN
10,6 uN
10,7 uN
10,8 uN
10,9 uN
10,10
3
˜∆10(µ)
A. Narayan (U. Utah) Emulators and surrogates
RBM accuracy
Does uN computed via RBM provide a good emulator for u? Depends on the
sampling.
Let u(xm) ∈ H. Suppose we choose
xn+1 = arg max
x∈D
un(x) − u(x) H
(This can be approximated without knowing u!)
Then,
u − uN L∞(D,H) σN (U),
where
U := u(x) x ∈ D ⊂ H,
σN (U) := inf
dim V =N
sup
v∈U
inf
vN ∈V
v − vN H.
[DeVore et al 2013], [Binev et al 2013]
A. Narayan (U. Utah) Emulators and surrogates
RBM accuracy
Surrogates for nontrivial problems can be constructed.
(−∆)s
u(z; x) = f(z; ν), (z, x) ∈ Ω × D
u(z; x) = 0, (z, x) ∈ ∂Ω × D
Parameters/variables are x = (s, ν).
4 2 4 6 8 10 12 14
10
-10
10-5
10
0
tion UN associated to problem (42), where µ = (s, ⌫).
1 100 200 309
M
10
2
103
10
4
10
5
tMsolves
s ∈ D1
M · tUN
toffline + M · tU10
1 100 200 309
M
10
2
103
10
4
10
5
tMsolves
s ∈ D2
M · tUN
toffline + M · tU10
Figure 3: The cumulative computation time for M queries of the full order model uN and the
RBM surrogate uN . On the left is for the case s 2 D1 with N = 7; on the right, s 2 D2 with
[Antil, Chen, AN 2018]
A. Narayan (U. Utah) Emulators and surrogates
Building emulators
Surrogate models can be enormously useful.
Linear approximations with non-adapted basis functions
“Easiest” to construct, with weakest accuracy guarantees.
Querying surrogate generally very fast.
Useful for analying large datasets
Nonlinear approximations with non-adapted basis functions
Harder to construct, but more general accuracy guarantees.
Querying surrogate still very fast.
Useful when data is limited.
Nonlinear approximations with adapted basis functions
Generally very hard to construct.
Very attractive accuracy bounds, when possible to certify
Depend heavily on data, model, and the transparency of the model.
A. Narayan (U. Utah) Emulators and surrogates
Building emulators
Surrogate models can be enormously useful.
Linear approximations with non-adapted basis functions
“Easiest” to construct, with weakest accuracy guarantees.
Querying surrogate generally very fast.
Useful for analying large datasets
Nonlinear approximations with non-adapted basis functions
Harder to construct, but more general accuracy guarantees.
Querying surrogate still very fast.
Useful when data is limited.
Nonlinear approximations with adapted basis functions
Generally very hard to construct.
Very attractive accuracy bounds, when possible to certify
Depend heavily on data, model, and the transparency of the model.
Challenges:
high dimensionality (d, P, or N)
adaptivity and hierarchical constructions
A. Narayan (U. Utah) Emulators and surrogates
mathematics of reduced order models
algorithms for approximation and
complexity reduction
computational statistics and data-driven
techniques
https://icerm.brown.edu/programs/sp-s20/
A. Narayan (U. Utah) Emulators and surrogates
References
Chkifa, Cohen, Migliorati, Nobile, Tempone, ”Discrete least squares polynomial approximation with
random evaluations – application to parametric and stochastic elliptic PDEs”, ESAIM: Mathematical
Modelling and Numerical Analysis, 49:3 (2015)
Cohen, Davenport, & Leviatan, ”On the Stability and Accuracy of Least Squares Approximations”,
Foundations of Computational Mathematics, 13:5 (2013)
Cohen & Migliorati, ”Optimal weighted least-squares methods”, arXiv:1608.00512 [math, stat]
Jakeman, Narayan, & Zhou, ”A Christoffel function weighted least squares algorithm for collocation
approximations”, Mathematics of Computation, 86:306 (2017)
Narayan, ”Computation of Induced Orthogonal Polynomial Distributions”, arXiv:1704.08465 [math]
(2017)
Narayan & Zhou, ”Stochastic Collocation on Unstructured Multivariate Meshes”, Communications in
Computational Physics, 18:1 (2015)
A. Narayan (U. Utah) Emulators and surrogates

More Related Content

What's hot

My data are incomplete and noisy: Information-reduction statistical methods f...
My data are incomplete and noisy: Information-reduction statistical methods f...My data are incomplete and noisy: Information-reduction statistical methods f...
My data are incomplete and noisy: Information-reduction statistical methods f...Umberto Picchini
 
. An introduction to machine learning and probabilistic ...
. An introduction to machine learning and probabilistic .... An introduction to machine learning and probabilistic ...
. An introduction to machine learning and probabilistic ...butest
 
A likelihood-free version of the stochastic approximation EM algorithm (SAEM)...
A likelihood-free version of the stochastic approximation EM algorithm (SAEM)...A likelihood-free version of the stochastic approximation EM algorithm (SAEM)...
A likelihood-free version of the stochastic approximation EM algorithm (SAEM)...Umberto Picchini
 
from model uncertainty to ABC
from model uncertainty to ABCfrom model uncertainty to ABC
from model uncertainty to ABCChristian Robert
 
Application of interpolation and finite difference
Application of interpolation and finite differenceApplication of interpolation and finite difference
Application of interpolation and finite differenceManthan Chavda
 
On the vexing dilemma of hypothesis testing and the predicted demise of the B...
On the vexing dilemma of hypothesis testing and the predicted demise of the B...On the vexing dilemma of hypothesis testing and the predicted demise of the B...
On the vexing dilemma of hypothesis testing and the predicted demise of the B...Christian Robert
 
Equational axioms for probability calculus and modelling of Likelihood ratio ...
Equational axioms for probability calculus and modelling of Likelihood ratio ...Equational axioms for probability calculus and modelling of Likelihood ratio ...
Equational axioms for probability calculus and modelling of Likelihood ratio ...Advanced-Concepts-Team
 
Thesis Presentation_Extreme Learning Machine_Nimai_SC14M045
Thesis Presentation_Extreme Learning Machine_Nimai_SC14M045Thesis Presentation_Extreme Learning Machine_Nimai_SC14M045
Thesis Presentation_Extreme Learning Machine_Nimai_SC14M045Nimai Chand Das Adhikari
 
random forests for ABC model choice and parameter estimation
random forests for ABC model choice and parameter estimationrandom forests for ABC model choice and parameter estimation
random forests for ABC model choice and parameter estimationChristian Robert
 
NBBC15, Reyjavik, June 08, 2015
NBBC15, Reyjavik, June 08, 2015NBBC15, Reyjavik, June 08, 2015
NBBC15, Reyjavik, June 08, 2015Christian Robert
 
CISEA 2019: ABC consistency and convergence
CISEA 2019: ABC consistency and convergenceCISEA 2019: ABC consistency and convergence
CISEA 2019: ABC consistency and convergenceChristian Robert
 
Harvard_University_-_Linear_Al
Harvard_University_-_Linear_AlHarvard_University_-_Linear_Al
Harvard_University_-_Linear_Alramiljayureta
 
Nature-Inspired Metaheuristic Algorithms for Optimization and Computational I...
Nature-Inspired Metaheuristic Algorithms for Optimization and Computational I...Nature-Inspired Metaheuristic Algorithms for Optimization and Computational I...
Nature-Inspired Metaheuristic Algorithms for Optimization and Computational I...Xin-She Yang
 
Laplace's Demon: seminar #1
Laplace's Demon: seminar #1Laplace's Demon: seminar #1
Laplace's Demon: seminar #1Christian Robert
 

What's hot (20)

My data are incomplete and noisy: Information-reduction statistical methods f...
My data are incomplete and noisy: Information-reduction statistical methods f...My data are incomplete and noisy: Information-reduction statistical methods f...
My data are incomplete and noisy: Information-reduction statistical methods f...
 
Estimating Space-Time Covariance from Finite Sample Sets
Estimating Space-Time Covariance from Finite Sample SetsEstimating Space-Time Covariance from Finite Sample Sets
Estimating Space-Time Covariance from Finite Sample Sets
 
. An introduction to machine learning and probabilistic ...
. An introduction to machine learning and probabilistic .... An introduction to machine learning and probabilistic ...
. An introduction to machine learning and probabilistic ...
 
A likelihood-free version of the stochastic approximation EM algorithm (SAEM)...
A likelihood-free version of the stochastic approximation EM algorithm (SAEM)...A likelihood-free version of the stochastic approximation EM algorithm (SAEM)...
A likelihood-free version of the stochastic approximation EM algorithm (SAEM)...
 
Intractable likelihoods
Intractable likelihoodsIntractable likelihoods
Intractable likelihoods
 
from model uncertainty to ABC
from model uncertainty to ABCfrom model uncertainty to ABC
from model uncertainty to ABC
 
Application of interpolation and finite difference
Application of interpolation and finite differenceApplication of interpolation and finite difference
Application of interpolation and finite difference
 
On the vexing dilemma of hypothesis testing and the predicted demise of the B...
On the vexing dilemma of hypothesis testing and the predicted demise of the B...On the vexing dilemma of hypothesis testing and the predicted demise of the B...
On the vexing dilemma of hypothesis testing and the predicted demise of the B...
 
Equational axioms for probability calculus and modelling of Likelihood ratio ...
Equational axioms for probability calculus and modelling of Likelihood ratio ...Equational axioms for probability calculus and modelling of Likelihood ratio ...
Equational axioms for probability calculus and modelling of Likelihood ratio ...
 
Thesis Presentation_Extreme Learning Machine_Nimai_SC14M045
Thesis Presentation_Extreme Learning Machine_Nimai_SC14M045Thesis Presentation_Extreme Learning Machine_Nimai_SC14M045
Thesis Presentation_Extreme Learning Machine_Nimai_SC14M045
 
Polynomial Matrix Decompositions
Polynomial Matrix DecompositionsPolynomial Matrix Decompositions
Polynomial Matrix Decompositions
 
ABC-Gibbs
ABC-GibbsABC-Gibbs
ABC-Gibbs
 
random forests for ABC model choice and parameter estimation
random forests for ABC model choice and parameter estimationrandom forests for ABC model choice and parameter estimation
random forests for ABC model choice and parameter estimation
 
NBBC15, Reyjavik, June 08, 2015
NBBC15, Reyjavik, June 08, 2015NBBC15, Reyjavik, June 08, 2015
NBBC15, Reyjavik, June 08, 2015
 
ABC-Gibbs
ABC-GibbsABC-Gibbs
ABC-Gibbs
 
CISEA 2019: ABC consistency and convergence
CISEA 2019: ABC consistency and convergenceCISEA 2019: ABC consistency and convergence
CISEA 2019: ABC consistency and convergence
 
Harvard_University_-_Linear_Al
Harvard_University_-_Linear_AlHarvard_University_-_Linear_Al
Harvard_University_-_Linear_Al
 
asymptotics of ABC
asymptotics of ABCasymptotics of ABC
asymptotics of ABC
 
Nature-Inspired Metaheuristic Algorithms for Optimization and Computational I...
Nature-Inspired Metaheuristic Algorithms for Optimization and Computational I...Nature-Inspired Metaheuristic Algorithms for Optimization and Computational I...
Nature-Inspired Metaheuristic Algorithms for Optimization and Computational I...
 
Laplace's Demon: seminar #1
Laplace's Demon: seminar #1Laplace's Demon: seminar #1
Laplace's Demon: seminar #1
 

Similar to MUMS Opening Workshop - Emulators for models and Complexity Reduction - Akil Narayan , August 21, 2018

Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...Valentin De Bortoli
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligencekeerthikaA8
 
Artificial intelligence.pptx
Artificial intelligence.pptxArtificial intelligence.pptx
Artificial intelligence.pptxkeerthikaA8
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligencekeerthikaA8
 
Moshe Guttmann's slides on eigenface
Moshe Guttmann's slides on eigenfaceMoshe Guttmann's slides on eigenface
Moshe Guttmann's slides on eigenfacewolf
 
Cheatsheet unsupervised-learning
Cheatsheet unsupervised-learningCheatsheet unsupervised-learning
Cheatsheet unsupervised-learningSteve Nouri
 
Cheatsheet supervised-learning
Cheatsheet supervised-learningCheatsheet supervised-learning
Cheatsheet supervised-learningSteve Nouri
 
Pattern learning and recognition on statistical manifolds: An information-geo...
Pattern learning and recognition on statistical manifolds: An information-geo...Pattern learning and recognition on statistical manifolds: An information-geo...
Pattern learning and recognition on statistical manifolds: An information-geo...Frank Nielsen
 
On non-negative unbiased estimators
On non-negative unbiased estimatorsOn non-negative unbiased estimators
On non-negative unbiased estimatorsPierre Jacob
 
Slides: Hypothesis testing, information divergence and computational geometry
Slides: Hypothesis testing, information divergence and computational geometrySlides: Hypothesis testing, information divergence and computational geometry
Slides: Hypothesis testing, information divergence and computational geometryFrank Nielsen
 
Monte Carlo Methods
Monte Carlo MethodsMonte Carlo Methods
Monte Carlo MethodsJames Bell
 
Master Thesis on the Mathematial Analysis of Neural Networks
Master Thesis on the Mathematial Analysis of Neural NetworksMaster Thesis on the Mathematial Analysis of Neural Networks
Master Thesis on the Mathematial Analysis of Neural NetworksAlina Leidinger
 
Radial Basis Function Interpolation
Radial Basis Function InterpolationRadial Basis Function Interpolation
Radial Basis Function InterpolationJesse Bettencourt
 
Nonlinear Manifolds in Computer Vision
Nonlinear Manifolds in Computer VisionNonlinear Manifolds in Computer Vision
Nonlinear Manifolds in Computer Visionzukun
 
Non Linear Equation
Non Linear EquationNon Linear Equation
Non Linear EquationMdAlAmin187
 

Similar to MUMS Opening Workshop - Emulators for models and Complexity Reduction - Akil Narayan , August 21, 2018 (20)

overviewPCA
overviewPCAoverviewPCA
overviewPCA
 
Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligence
 
Artificial intelligence.pptx
Artificial intelligence.pptxArtificial intelligence.pptx
Artificial intelligence.pptx
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligence
 
Moshe Guttmann's slides on eigenface
Moshe Guttmann's slides on eigenfaceMoshe Guttmann's slides on eigenface
Moshe Guttmann's slides on eigenface
 
Cheatsheet unsupervised-learning
Cheatsheet unsupervised-learningCheatsheet unsupervised-learning
Cheatsheet unsupervised-learning
 
Cheatsheet supervised-learning
Cheatsheet supervised-learningCheatsheet supervised-learning
Cheatsheet supervised-learning
 
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
 
Pattern learning and recognition on statistical manifolds: An information-geo...
Pattern learning and recognition on statistical manifolds: An information-geo...Pattern learning and recognition on statistical manifolds: An information-geo...
Pattern learning and recognition on statistical manifolds: An information-geo...
 
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
 
On non-negative unbiased estimators
On non-negative unbiased estimatorsOn non-negative unbiased estimators
On non-negative unbiased estimators
 
CLIM Fall 2017 Course: Statistics for Climate Research, Climate Informatics -...
CLIM Fall 2017 Course: Statistics for Climate Research, Climate Informatics -...CLIM Fall 2017 Course: Statistics for Climate Research, Climate Informatics -...
CLIM Fall 2017 Course: Statistics for Climate Research, Climate Informatics -...
 
Slides: Hypothesis testing, information divergence and computational geometry
Slides: Hypothesis testing, information divergence and computational geometrySlides: Hypothesis testing, information divergence and computational geometry
Slides: Hypothesis testing, information divergence and computational geometry
 
Monte Carlo Methods
Monte Carlo MethodsMonte Carlo Methods
Monte Carlo Methods
 
Master Thesis on the Mathematial Analysis of Neural Networks
Master Thesis on the Mathematial Analysis of Neural NetworksMaster Thesis on the Mathematial Analysis of Neural Networks
Master Thesis on the Mathematial Analysis of Neural Networks
 
Radial Basis Function Interpolation
Radial Basis Function InterpolationRadial Basis Function Interpolation
Radial Basis Function Interpolation
 
Nested sampling
Nested samplingNested sampling
Nested sampling
 
Nonlinear Manifolds in Computer Vision
Nonlinear Manifolds in Computer VisionNonlinear Manifolds in Computer Vision
Nonlinear Manifolds in Computer Vision
 
Non Linear Equation
Non Linear EquationNon Linear Equation
Non Linear Equation
 

More from The Statistical and Applied Mathematical Sciences Institute

More from The Statistical and Applied Mathematical Sciences Institute (20)

Causal Inference Opening Workshop - Latent Variable Models, Causal Inference,...
Causal Inference Opening Workshop - Latent Variable Models, Causal Inference,...Causal Inference Opening Workshop - Latent Variable Models, Causal Inference,...
Causal Inference Opening Workshop - Latent Variable Models, Causal Inference,...
 
2019 Fall Series: Special Guest Lecture - 0-1 Phase Transitions in High Dimen...
2019 Fall Series: Special Guest Lecture - 0-1 Phase Transitions in High Dimen...2019 Fall Series: Special Guest Lecture - 0-1 Phase Transitions in High Dimen...
2019 Fall Series: Special Guest Lecture - 0-1 Phase Transitions in High Dimen...
 
Causal Inference Opening Workshop - Causal Discovery in Neuroimaging Data - F...
Causal Inference Opening Workshop - Causal Discovery in Neuroimaging Data - F...Causal Inference Opening Workshop - Causal Discovery in Neuroimaging Data - F...
Causal Inference Opening Workshop - Causal Discovery in Neuroimaging Data - F...
 
Causal Inference Opening Workshop - Smooth Extensions to BART for Heterogeneo...
Causal Inference Opening Workshop - Smooth Extensions to BART for Heterogeneo...Causal Inference Opening Workshop - Smooth Extensions to BART for Heterogeneo...
Causal Inference Opening Workshop - Smooth Extensions to BART for Heterogeneo...
 
Causal Inference Opening Workshop - A Bracketing Relationship between Differe...
Causal Inference Opening Workshop - A Bracketing Relationship between Differe...Causal Inference Opening Workshop - A Bracketing Relationship between Differe...
Causal Inference Opening Workshop - A Bracketing Relationship between Differe...
 
Causal Inference Opening Workshop - Testing Weak Nulls in Matched Observation...
Causal Inference Opening Workshop - Testing Weak Nulls in Matched Observation...Causal Inference Opening Workshop - Testing Weak Nulls in Matched Observation...
Causal Inference Opening Workshop - Testing Weak Nulls in Matched Observation...
 
Causal Inference Opening Workshop - Difference-in-differences: more than meet...
Causal Inference Opening Workshop - Difference-in-differences: more than meet...Causal Inference Opening Workshop - Difference-in-differences: more than meet...
Causal Inference Opening Workshop - Difference-in-differences: more than meet...
 
Causal Inference Opening Workshop - New Statistical Learning Methods for Esti...
Causal Inference Opening Workshop - New Statistical Learning Methods for Esti...Causal Inference Opening Workshop - New Statistical Learning Methods for Esti...
Causal Inference Opening Workshop - New Statistical Learning Methods for Esti...
 
Causal Inference Opening Workshop - Bipartite Causal Inference with Interfere...
Causal Inference Opening Workshop - Bipartite Causal Inference with Interfere...Causal Inference Opening Workshop - Bipartite Causal Inference with Interfere...
Causal Inference Opening Workshop - Bipartite Causal Inference with Interfere...
 
Causal Inference Opening Workshop - Bridging the Gap Between Causal Literatur...
Causal Inference Opening Workshop - Bridging the Gap Between Causal Literatur...Causal Inference Opening Workshop - Bridging the Gap Between Causal Literatur...
Causal Inference Opening Workshop - Bridging the Gap Between Causal Literatur...
 
Causal Inference Opening Workshop - Some Applications of Reinforcement Learni...
Causal Inference Opening Workshop - Some Applications of Reinforcement Learni...Causal Inference Opening Workshop - Some Applications of Reinforcement Learni...
Causal Inference Opening Workshop - Some Applications of Reinforcement Learni...
 
Causal Inference Opening Workshop - Bracketing Bounds for Differences-in-Diff...
Causal Inference Opening Workshop - Bracketing Bounds for Differences-in-Diff...Causal Inference Opening Workshop - Bracketing Bounds for Differences-in-Diff...
Causal Inference Opening Workshop - Bracketing Bounds for Differences-in-Diff...
 
Causal Inference Opening Workshop - Assisting the Impact of State Polcies: Br...
Causal Inference Opening Workshop - Assisting the Impact of State Polcies: Br...Causal Inference Opening Workshop - Assisting the Impact of State Polcies: Br...
Causal Inference Opening Workshop - Assisting the Impact of State Polcies: Br...
 
Causal Inference Opening Workshop - Experimenting in Equilibrium - Stefan Wag...
Causal Inference Opening Workshop - Experimenting in Equilibrium - Stefan Wag...Causal Inference Opening Workshop - Experimenting in Equilibrium - Stefan Wag...
Causal Inference Opening Workshop - Experimenting in Equilibrium - Stefan Wag...
 
Causal Inference Opening Workshop - Targeted Learning for Causal Inference Ba...
Causal Inference Opening Workshop - Targeted Learning for Causal Inference Ba...Causal Inference Opening Workshop - Targeted Learning for Causal Inference Ba...
Causal Inference Opening Workshop - Targeted Learning for Causal Inference Ba...
 
Causal Inference Opening Workshop - Bayesian Nonparametric Models for Treatme...
Causal Inference Opening Workshop - Bayesian Nonparametric Models for Treatme...Causal Inference Opening Workshop - Bayesian Nonparametric Models for Treatme...
Causal Inference Opening Workshop - Bayesian Nonparametric Models for Treatme...
 
2019 Fall Series: Special Guest Lecture - Adversarial Risk Analysis of the Ge...
2019 Fall Series: Special Guest Lecture - Adversarial Risk Analysis of the Ge...2019 Fall Series: Special Guest Lecture - Adversarial Risk Analysis of the Ge...
2019 Fall Series: Special Guest Lecture - Adversarial Risk Analysis of the Ge...
 
2019 Fall Series: Professional Development, Writing Academic Papers…What Work...
2019 Fall Series: Professional Development, Writing Academic Papers…What Work...2019 Fall Series: Professional Development, Writing Academic Papers…What Work...
2019 Fall Series: Professional Development, Writing Academic Papers…What Work...
 
2019 GDRR: Blockchain Data Analytics - Machine Learning in/for Blockchain: Fu...
2019 GDRR: Blockchain Data Analytics - Machine Learning in/for Blockchain: Fu...2019 GDRR: Blockchain Data Analytics - Machine Learning in/for Blockchain: Fu...
2019 GDRR: Blockchain Data Analytics - Machine Learning in/for Blockchain: Fu...
 
2019 GDRR: Blockchain Data Analytics - QuTrack: Model Life Cycle Management f...
2019 GDRR: Blockchain Data Analytics - QuTrack: Model Life Cycle Management f...2019 GDRR: Blockchain Data Analytics - QuTrack: Model Life Cycle Management f...
2019 GDRR: Blockchain Data Analytics - QuTrack: Model Life Cycle Management f...
 

Recently uploaded

Introduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher EducationIntroduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher Educationpboyjonauth
 
Proudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptxProudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptxthorishapillay1
 
Incoming and Outgoing Shipments in 1 STEP Using Odoo 17
Incoming and Outgoing Shipments in 1 STEP Using Odoo 17Incoming and Outgoing Shipments in 1 STEP Using Odoo 17
Incoming and Outgoing Shipments in 1 STEP Using Odoo 17Celine George
 
Pharmacognosy Flower 3. Compositae 2023.pdf
Pharmacognosy Flower 3. Compositae 2023.pdfPharmacognosy Flower 3. Compositae 2023.pdf
Pharmacognosy Flower 3. Compositae 2023.pdfMahmoud M. Sallam
 
18-04-UA_REPORT_MEDIALITERAСY_INDEX-DM_23-1-final-eng.pdf
18-04-UA_REPORT_MEDIALITERAСY_INDEX-DM_23-1-final-eng.pdf18-04-UA_REPORT_MEDIALITERAСY_INDEX-DM_23-1-final-eng.pdf
18-04-UA_REPORT_MEDIALITERAСY_INDEX-DM_23-1-final-eng.pdfssuser54595a
 
Alper Gobel In Media Res Media Component
Alper Gobel In Media Res Media ComponentAlper Gobel In Media Res Media Component
Alper Gobel In Media Res Media ComponentInMediaRes1
 
“Oh GOSH! Reflecting on Hackteria's Collaborative Practices in a Global Do-It...
“Oh GOSH! Reflecting on Hackteria's Collaborative Practices in a Global Do-It...“Oh GOSH! Reflecting on Hackteria's Collaborative Practices in a Global Do-It...
“Oh GOSH! Reflecting on Hackteria's Collaborative Practices in a Global Do-It...Marc Dusseiller Dusjagr
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdfSoniaTolstoy
 
Solving Puzzles Benefits Everyone (English).pptx
Solving Puzzles Benefits Everyone (English).pptxSolving Puzzles Benefits Everyone (English).pptx
Solving Puzzles Benefits Everyone (English).pptxOH TEIK BIN
 
Enzyme, Pharmaceutical Aids, Miscellaneous Last Part of Chapter no 5th.pdf
Enzyme, Pharmaceutical Aids, Miscellaneous Last Part of Chapter no 5th.pdfEnzyme, Pharmaceutical Aids, Miscellaneous Last Part of Chapter no 5th.pdf
Enzyme, Pharmaceutical Aids, Miscellaneous Last Part of Chapter no 5th.pdfSumit Tiwari
 
Organic Name Reactions for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions  for the students and aspirants of Chemistry12th.pptxOrganic Name Reactions  for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions for the students and aspirants of Chemistry12th.pptxVS Mahajan Coaching Centre
 
History Class XII Ch. 3 Kinship, Caste and Class (1).pptx
History Class XII Ch. 3 Kinship, Caste and Class (1).pptxHistory Class XII Ch. 3 Kinship, Caste and Class (1).pptx
History Class XII Ch. 3 Kinship, Caste and Class (1).pptxsocialsciencegdgrohi
 
ENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptx
ENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptxENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptx
ENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptxAnaBeatriceAblay2
 
Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)eniolaolutunde
 
_Math 4-Q4 Week 5.pptx Steps in Collecting Data
_Math 4-Q4 Week 5.pptx Steps in Collecting Data_Math 4-Q4 Week 5.pptx Steps in Collecting Data
_Math 4-Q4 Week 5.pptx Steps in Collecting DataJhengPantaleon
 
POINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptx
POINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptxPOINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptx
POINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptxSayali Powar
 

Recently uploaded (20)

Introduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher EducationIntroduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher Education
 
Proudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptxProudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptx
 
Incoming and Outgoing Shipments in 1 STEP Using Odoo 17
Incoming and Outgoing Shipments in 1 STEP Using Odoo 17Incoming and Outgoing Shipments in 1 STEP Using Odoo 17
Incoming and Outgoing Shipments in 1 STEP Using Odoo 17
 
Pharmacognosy Flower 3. Compositae 2023.pdf
Pharmacognosy Flower 3. Compositae 2023.pdfPharmacognosy Flower 3. Compositae 2023.pdf
Pharmacognosy Flower 3. Compositae 2023.pdf
 
18-04-UA_REPORT_MEDIALITERAСY_INDEX-DM_23-1-final-eng.pdf
18-04-UA_REPORT_MEDIALITERAСY_INDEX-DM_23-1-final-eng.pdf18-04-UA_REPORT_MEDIALITERAСY_INDEX-DM_23-1-final-eng.pdf
18-04-UA_REPORT_MEDIALITERAСY_INDEX-DM_23-1-final-eng.pdf
 
Model Call Girl in Tilak Nagar Delhi reach out to us at 🔝9953056974🔝
Model Call Girl in Tilak Nagar Delhi reach out to us at 🔝9953056974🔝Model Call Girl in Tilak Nagar Delhi reach out to us at 🔝9953056974🔝
Model Call Girl in Tilak Nagar Delhi reach out to us at 🔝9953056974🔝
 
Alper Gobel In Media Res Media Component
Alper Gobel In Media Res Media ComponentAlper Gobel In Media Res Media Component
Alper Gobel In Media Res Media Component
 
“Oh GOSH! Reflecting on Hackteria's Collaborative Practices in a Global Do-It...
“Oh GOSH! Reflecting on Hackteria's Collaborative Practices in a Global Do-It...“Oh GOSH! Reflecting on Hackteria's Collaborative Practices in a Global Do-It...
“Oh GOSH! Reflecting on Hackteria's Collaborative Practices in a Global Do-It...
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
 
Solving Puzzles Benefits Everyone (English).pptx
Solving Puzzles Benefits Everyone (English).pptxSolving Puzzles Benefits Everyone (English).pptx
Solving Puzzles Benefits Everyone (English).pptx
 
Staff of Color (SOC) Retention Efforts DDSD
Staff of Color (SOC) Retention Efforts DDSDStaff of Color (SOC) Retention Efforts DDSD
Staff of Color (SOC) Retention Efforts DDSD
 
Enzyme, Pharmaceutical Aids, Miscellaneous Last Part of Chapter no 5th.pdf
Enzyme, Pharmaceutical Aids, Miscellaneous Last Part of Chapter no 5th.pdfEnzyme, Pharmaceutical Aids, Miscellaneous Last Part of Chapter no 5th.pdf
Enzyme, Pharmaceutical Aids, Miscellaneous Last Part of Chapter no 5th.pdf
 
Organic Name Reactions for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions  for the students and aspirants of Chemistry12th.pptxOrganic Name Reactions  for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions for the students and aspirants of Chemistry12th.pptx
 
TataKelola dan KamSiber Kecerdasan Buatan v022.pdf
TataKelola dan KamSiber Kecerdasan Buatan v022.pdfTataKelola dan KamSiber Kecerdasan Buatan v022.pdf
TataKelola dan KamSiber Kecerdasan Buatan v022.pdf
 
History Class XII Ch. 3 Kinship, Caste and Class (1).pptx
History Class XII Ch. 3 Kinship, Caste and Class (1).pptxHistory Class XII Ch. 3 Kinship, Caste and Class (1).pptx
History Class XII Ch. 3 Kinship, Caste and Class (1).pptx
 
9953330565 Low Rate Call Girls In Rohini Delhi NCR
9953330565 Low Rate Call Girls In Rohini  Delhi NCR9953330565 Low Rate Call Girls In Rohini  Delhi NCR
9953330565 Low Rate Call Girls In Rohini Delhi NCR
 
ENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptx
ENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptxENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptx
ENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptx
 
Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)
 
_Math 4-Q4 Week 5.pptx Steps in Collecting Data
_Math 4-Q4 Week 5.pptx Steps in Collecting Data_Math 4-Q4 Week 5.pptx Steps in Collecting Data
_Math 4-Q4 Week 5.pptx Steps in Collecting Data
 
POINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptx
POINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptxPOINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptx
POINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptx
 

MUMS Opening Workshop - Emulators for models and Complexity Reduction - Akil Narayan , August 21, 2018

  • 1. Emulators for models and complexity reduction Akil Narayan1 1 Department of Mathematics, and Scientific Computing and Imaging (SCI) Institute University of Utah August 2018 SAMSI MUMS opening workshop A. Narayan (U. Utah) Emulators and surrogates
  • 2. Models and emulators y = u(x) + ε x ∈ D ⊆ Rd y ∈ RP x y The parameters/factors x govern the bulk behavior of the response u The noise or error ε can account for model discrepancy The observable y can be deterministic or stochastic A. Narayan (U. Utah) Emulators and surrogates
  • 3. Models and emulators y = u(x) + ε x ∈ D ⊆ Rd y ∈ RP x y Available data: noisy measurements, y, abstractly treated as samples at specific values of x. Emulators are generally built to be consistent with data. Their purpose can be to extrapolate/interpolate data accelerate queries of the model analyze for variances, screening, sensitivity, etc. A. Narayan (U. Utah) Emulators and surrogates
  • 4. Models and emulators y = u(x) + ε x ∈ D ⊆ Rd y ∈ RP x y I will primarily discuss emulator constructions from applied mathematics/scientific computing. We are interested in things like stability, accuracy, consistency, etc. Take-home point: experimental design is critical in building good emulators. A. Narayan (U. Utah) Emulators and surrogates
  • 5. Building emulators Many mathematical emulator models have the form u(x) ≈ uN (x) := N n=1 cnφn(x), Information about y: sample data (xm, ym), m = 1, . . . , M. Two general types of approximations: linear approximations: uN is linear in the data. φn(·) are prescribed a priori, {ym} → {cn} is linear nonlinear approximations: uN is nonlinearly dependent on data computation of cn may be nonlinear identification of φn may depend on data A. Narayan (U. Utah) Emulators and surrogates
  • 6. Building emulators Many mathematical emulator models have the form u(x) ≈ uN (x) := N n=1 cnφn(x), Information about y: sample data (xm, ym), m = 1, . . . , M. Two general types of approximations: linear approximations: uN is linear in the data. φn(·) are prescribed a priori, {ym} → {cn} is linear nonlinear approximations: uN is nonlinearly dependent on data computation of cn may be nonlinear identification of φn may depend on data The form of φn does not generally dictate linear/nonlinearity. Some linear approximations: interpolation quadrature least-squares Some nonlinear approximations: radial basis/kernel approximations non-quadratic regularized approximation proper orthogonal decomposition A. Narayan (U. Utah) Emulators and surrogates
  • 7. Building emulators Many mathematical emulator models have the form u(x) ≈ uN (x) := N n=1 cnφn(x), Information about y: sample data (xm, ym), m = 1, . . . , M. Two general types of approximations: linear approximations: uN is linear in the data. φn(·) are prescribed a priori, {ym} → {cn} is linear nonlinear approximations: uN is nonlinearly dependent on data computation of cn may be nonlinear identification of φn may depend on data Example: If M ≥ N, cj coefficients computable via least-squares      y1 y2 ... yM      = y ≈ Ac =      φ1(x1) φ2(x1) · · · φN (x1) φ1(x2) φ2(x2) · · · φN (x2) ... ... ... ... φ1(xM ) φ2(xM ) · · · φN (xM )           c1 c2 ... cN      A. Narayan (U. Utah) Emulators and surrogates
  • 8. Emulators as model reduction Emulators are built in the hope that x → u(x) is a map of low complexity. If true, and an efficient model to capture this complexity is discoverable, then u(x) ≈ uN (x) = N n=1 cnφn(x), V := span{φ1, . . . , φN } can be achieved with “small” N. A. Narayan (U. Utah) Emulators and surrogates
  • 9. Emulators as model reduction Emulators are built in the hope that x → u(x) is a map of low complexity. If true, and an efficient model to capture this complexity is discoverable, then u(x) ≈ uN (x) = N n=1 cnφn(x), V := span{φ1, . . . , φN } can be achieved with “small” N. Identify V Efficiently construct uN from V Neither of these is particularly easy in general. Anyway, scientific models are complex, is this even feasible with reasonable N? A. Narayan (U. Utah) Emulators and surrogates
  • 10. An explicit example Example: Consider the solution u(z; x) to the parameterized PDE: − z · (a(z; x) zu(z; x)) = f(z), (z, x) ∈ Ω × D, u(z; x) = 0, (z, x) ∈ ∂Ω × D. For each x, u(·; x) ∈ H = H1 (Ω). Let the diffusion coefficient be given by a(z; x) = ∞ j=1 xjψj(z). A. Narayan (U. Utah) Emulators and surrogates
  • 11. An explicit example Example: Consider the solution u(z; x) to the parameterized PDE: − z · (a(z; x) zu(z; x)) = f(z), (z, x) ∈ Ω × D, u(z; x) = 0, (z, x) ∈ ∂Ω × D. For each x, u(·; x) ∈ H = H1 (Ω). Let the diffusion coefficient be given by a(z; x) = ∞ j=1 xjψj(z). If x = (x1, . . .) ∈ D = [−1, 1]∞ , and there is some p ≤ 1 such that ∞ j=1 ψj p L∞(Ω) < ∞, then an emulator uN can be constructed such that u − uN L2(D,H) N−r , r = 1 p − 1 2 . [Cohen, DeVore, Schwab 2010] A. Narayan (U. Utah) Emulators and surrogates
  • 12. Adapted vs linear An approximation to u: u ≈ uN (z; x) = N n=1 cn(z)φn(x), V := span{φ1, . . . , φN } Non-adapted approximation: With V chosen, construct uN so that u − uN L2(D,RP ) inf v∈V u − v L2(D,RP ) The main task is to compute uN from a given V . A. Narayan (U. Utah) Emulators and surrogates
  • 13. Adapted vs linear An approximation to u: u ≈ uN (z; x) = N n=1 cn(z)φn(x), V := span{φ1, . . . , φN } Non-adapted approximation: With V chosen, construct uN so that u − uN L2(D,RP ) inf v∈V u − v L2(D,RP ) The main task is to compute uN from a given V . Adapated approximation: Find V and uN so that u(x) − uN (x) RP is “small” for all x ∈ D A. Narayan (U. Utah) Emulators and surrogates
  • 14. Adapted vs linear An approximation to u: u ≈ uN (z; x) = N n=1 cn(z)φn(x), V := span{φ1, . . . , φN } Non-adapted approximation: With V chosen, construct uN so that u − uN L2(D,RP ) inf v∈V u − v L2(D,RP ) The main task is to compute uN from a given V . Adapated approximation: Find V and uN so that u(x) − uN (x) RP is “small” for all x ∈ D Adapted approximations are always nonlinear. Non-adapted approximations can be linear. A. Narayan (U. Utah) Emulators and surrogates
  • 15. Emulators and sampling/experimental design y = u + u ≈ uN = N n=1 cnφn(x) ∈ V, {(xm, ym)} M m=1 −→ {cn} N n=1 Desiderata: u − uN B small for a normed vector space B M of “reasonable” size Accuracy, both in identification of V and in computation of uN depends largely on sample design, i.e., the choice of x1, . . . , xM . A. Narayan (U. Utah) Emulators and surrogates
  • 16. Emulators and sampling/experimental design y = u + u ≈ uN = N n=1 cnφn(x) ∈ V, {(xm, ym)} M m=1 −→ {cn} N n=1 Desiderata: u − uN B small for a normed vector space B M of “reasonable” size Accuracy, both in identification of V and in computation of uN depends largely on sample design, i.e., the choice of x1, . . . , xM . Good sample design can minimize required data size M Intelligent sampling enables efficient emulator construction A. Narayan (U. Utah) Emulators and surrogates
  • 17. Summary of methods We’ll see how sampling design affects approximation statements for three strategies: Discrete least-squares: linear approximation, M ≥ N Compressive sampling: nonlinear approximation, M N Reduced order modeling: nonlinear approximation, N ∼ M = O(1) I’ll discuss optimal mathematical statements one can make, taking the form u − uN B KN × (Best approx error) + if M ≥ KM . I will focus on the role that sampling plays in these techniques. A. Narayan (U. Utah) Emulators and surrogates
  • 18. Summary of methods We’ll see how sampling design affects approximation statements for three strategies: Discrete least-squares: linear approximation, M ≥ N Compressive sampling: nonlinear approximation, M N Reduced order modeling: nonlinear approximation, N ∼ M = O(1) I’ll discuss optimal mathematical statements one can make, taking the form u − uN B KN × (Best approx error) + if M ≥ KM . I will focus on the role that sampling plays in these techniques. Warning: There are entire sub-fields of applied math and statistics concerning sampling that I will ignore. A. Narayan (U. Utah) Emulators and surrogates
  • 19. Summary of methods We’ll see how sampling design affects approximation statements for three strategies: Discrete least-squares: linear approximation, M ≥ N Compressive sampling: nonlinear approximation, M N Reduced order modeling: nonlinear approximation, N ∼ M = O(1) I’ll discuss optimal mathematical statements one can make, taking the form u − uN B KN × (Best approx error) + if M ≥ KM . I will focus on the role that sampling plays in these techniques. Warning: There are entire sub-fields of applied math and statistics concerning sampling that I will ignore. (Because they’re not directly relevant to the message.) A. Narayan (U. Utah) Emulators and surrogates
  • 20. Part I: Linear approximation Discrete least squares Non-adapted basis functions, linear approximation construction procedure A. Narayan (U. Utah) Emulators and surrogates
  • 21. An aside – polynomials and PCE u(x) + = y ≈ uN (x) = N n=1 cnφn(x), x ∈ D ⊆ Rd y ∈ R c ∈ RN Non-adapted approximation: the φn are a priori chosen. We often choose polynomials. (Cf. PCE) Why? A. Narayan (U. Utah) Emulators and surrogates
  • 22. An aside – polynomials and PCE u(x) + = y ≈ uN (x) = N n=1 cnφn(x), x ∈ D ⊆ Rd y ∈ R c ∈ RN Non-adapted approximation: the φn are a priori chosen. We often choose polynomials. (Cf. PCE) Why? Polynomials are easy to compute with/evaluate Polynomial expansions are (reasonably) easy to manipulate, multiply, differentiate, etc. Polynomials provide best approximation numbers that behave optimally: inf N=dim P =dim P d k P ⊂Hs sup f∈Hs f Hs =1 inf p∈P f − p Hs ∼ N−s/d sup f∈Hs f Hs =1 inf p∈P d k N=dim P d k f − p Hs N−s/d [Pinkus 1985] A. Narayan (U. Utah) Emulators and surrogates
  • 23. Mathematical preliminaries u(x) + = y ≈ uN (x) = N n=1 cnφn(x), x ∈ D ⊆ Rd y ∈ R c ∈ RN Least-squares problem is approximation of the form (c∗ 1, . . . , c∗ N ) T = c∗ = arg min c∈RN M m=1 [uN (xm) − ym] 2 A. Narayan (U. Utah) Emulators and surrogates
  • 24. Mathematical preliminaries u(x) + = y ≈ uN (x) = N n=1 cnφn(x), x ∈ D ⊆ Rd y ∈ R c ∈ RN Least-squares problem is approximation of the form (c∗ 1, . . . , c∗ N ) T = c∗ = arg min c∈RN M m=1 [uN (xm) − ym] 2 Let V := span {φ1, . . . , φN }. Least-squares is, equivalently, v∗ = arg min v∈V M m=1 (v(xm) − ym) 2 V is an a priori space of functions. What is the “best” approximation we can hope for? A. Narayan (U. Utah) Emulators and surrogates
  • 25. Mathematical preliminaries Given a probability measure µ on D, approximation will take place in an L2 space: g, h µ := D g(x)h(x)dµ(x), L2 µ (D) := g : D → R g µ < ∞ The best approximation error to u from the subspace V is σV (u) := inf v∈V u − v µ A. Narayan (U. Utah) Emulators and surrogates
  • 26. Mathematical preliminaries Given a probability measure µ on D, approximation will take place in an L2 space: g, h µ := D g(x)h(x)dµ(x), L2 µ (D) := g : D → R g µ < ∞ The best approximation error to u from the subspace V is σV (u) := inf v∈V u − v µ Randomized sampling: xm sampled iid from µ, and no noise, ym = u(xm), uN = arg min v∈V M m=1 (v(xm) − ym) 2 Law of large numbers: M ↑ ∞ ⇒ uN − u µ → σV (u). A. Narayan (U. Utah) Emulators and surrogates
  • 27. “Standard” Monte Carlo Approximate a function u(x) = exp −ω x − 1 π 2 , x ∈ [−1, 1], = 0, with µ uniform on [−1, 1], from the space of potential surrogates V = span 1, . . . , xN−1 Data xm sampled iid from µ Convergence observed, but slow Why does this happen, and can we fix it? 50 100 150 200 250 300 10−5 10−3 10−1 101 103 105 107 M Mean-squareerror D = [−1, 1], N = 50 Optimal error MC A. Narayan (U. Utah) Emulators and surrogates
  • 28. “Standard” Monte Carlo Approximate a function u(x) = exp −ω x − 1 π 2 , x ∈ [−1, 1], = 0, with µ uniform on [−1, 1], from the space of potential surrogates V = span 1, . . . , xN−1 Data xm sampled iid from µ Convergence observed, but slow Why does this happen, and can we fix it? 50 100 150 200 250 300 10−5 10−3 10−1 101 103 105 107 M Mean-squareerror D = [−1, 1], N = 50 Optimal error MC Sampling from a standard distribution is frequently suboptimal A. Narayan (U. Utah) Emulators and surrogates
  • 29. Convergence results Proximity to the optimal solution is guaranteed with enough samples. Define Kµ(V ) := sup x∈D sup v∈V {0} |v(x)|2 v 2 µ If x1, . . . , xM are sampled iid from µ, then M log M ≥ 2 + 2r log(e/2) Kµ(V ) guarantees that, with probability ≥ 1 − 2M−r , E u − uN 2 µ ≤ 1 + 2 1 − log 2 (1 + r) log M σV (u)2 + 8U2 M−r , where U = supx∈D |u(x)|, and uN = arg min v∈V M m=1 (v(xm) − ym) 2 [Cohen, Davenport, Leviatan 2013] A. Narayan (U. Utah) Emulators and surrogates
  • 30. Randomized sampling – Monte Carlo M log M ≥ 2 + 2r log(e/2) Kµ(V ), Kµ(V ) = sup x∈D sup v∈V {0} |v(x)|2 v 2 µ The smallest (best) value of Kµ(V ) is N. A. Narayan (U. Utah) Emulators and surrogates
  • 31. Randomized sampling – Monte Carlo M log M ≥ 2 + 2r log(e/2) Kµ(V ), Kµ(V ) = sup x∈D sup v∈V {0} |v(x)|2 v 2 µ The smallest (best) value of Kµ(V ) is N. Example: Linear models, N = d + 1 φ1(x) = 1, φj+1(x) = xj, j = 1, . . . , d Let µ be the standard Gaussian measure over D = Rd Then Kµ(V ) = ∞. Analysis suggests this is a pretty bad sampling design, but in practice it’s fine. A. Narayan (U. Utah) Emulators and surrogates
  • 32. Randomized sampling – Monte Carlo M log M ≥ 2 + 2r log(e/2) Kµ(V ), Kµ(V ) = sup x∈D sup v∈V {0} |v(x)|2 v 2 µ The smallest (best) value of Kµ(V ) is N. Example: Linear models, N = d + 1 φ1(x) = 1, φj+1(x) = xj, j = 1, . . . , d Let µ be the standard Gaussian measure over D = Rd Then Kµ(V ) = ∞. Analysis suggests this is a pretty bad sampling design, but in practice it’s fine. In the previous example, Kµ(V ) ∼ N2 . In practice, Kµ(V ) depends exponentially on d. The ideal case: Kµ(V ) ∼ N. To accomplish this, use biased sampling. A. Narayan (U. Utah) Emulators and surrogates
  • 33. Randomized sampling – weighted methods Lesson: sampling xm ∼ µ is usually not optimal, and sometimes terrible. Standard least-squares: arg min c Ac − y 2 Weighted least-squares: arg min c Ac − y 2,w = arg min c √ W Ac − √ W y 2 where W = diag(w1, . . . , wM ) contains positive weights wj. A. Narayan (U. Utah) Emulators and surrogates
  • 34. Randomized sampling – optimality We can entirely circumvent the Kµ(V ) problem by changing sampling measures. Assume φ1, . . . φN is an L2 µ-orthonormal basis for V . Generate x1, . . . , xM iid from µV , where dµV (x) = 1 N N n=1 φ2 n(x)dµ(x). Use weights wm = dµ dµV (xm) = N N n=1 φ2 n(xm) . Our weighted least-squares estimator is defined by c∗ = arg min c Ac − y 2,w The measure µV is called the induced distribution for V . A. Narayan (U. Utah) Emulators and surrogates
  • 35. Randomized sampling – optimality Let x1, . . . , xM ∼ µV , with uN (x) = N n=1 c∗ nφn(x) computed via c∗ = arg min c Ac − y 2,w Then M log M ≥ 2 + 2r log(e/2) N guarantees that, with probability ≥ 1 − 2M−r , E u − uN 2 µ ≤ 1 + 2 1 − log 2 (1 + r) log M σV (µ)2 + 8U2 M−r [Cohen, Migliorati 2017] Note: This M/N dependence is essentially optimal. A. Narayan (U. Utah) Emulators and surrogates
  • 36. The induced distribution The induced distribution µV can be substantially different from µ. x ∈ D = R2 , dµ(x) ∝ exp(− x 2 2), x = (x(1) , x(2) ), V = span x(1) α1 x(2) α2 (α1 + 1)(α2 + 1) ≤ 26 −6 −5 −4 −3 −2 −1 0 1 2 3 4 5 6 −6 −5 −4 −3 −2 −1 0 1 2 3 4 5 6 x(1) x(2) Samples from µ −6 −5 −4 −3 −2 −1 0 1 2 3 4 5 6 −6 −5 −4 −3 −2 −1 0 1 2 3 4 5 6 x(1) x(2) Samples from µV Under certain conditions, can sample from this distribution very efficiently, in particular with linear complexity in d. [AN 2017] A. Narayan (U. Utah) Emulators and surrogates
  • 37. Randomized sampling – examples This analysis tends to give accurate estimates 10 20 30 40 50 10−1 100 101 M L2 µerror N = 5 Optimal error = ε µ µ µV 50 100 150 200 250 300 10−5 10−3 10−1 101 103 105 107 M L2 µerror N = 45 A. Narayan (U. Utah) Emulators and surrogates
  • 38. Randomized sampling – examples This analysis tends to give accurate estimates 10 20 30 40 50 10−1 100 101 M L2 µerror N = 5 Optimal error = ε µ µ µV 50 100 150 200 250 300 10−5 10−3 10−1 101 103 105 107 M L2 µerror N = 45 Moral of the story: randomized sampling according to µ is generally bad randomized sampling according to µV is generally good Intelligent sampling allows efficient, near-optimal computation of emulators. A. Narayan (U. Utah) Emulators and surrogates
  • 39. Odds and ends Robust and accurate least-squares emulators for linear approximations can be built with biased sampling. Estimates are optimal: M N implies u − uN µ σV (u). Estimates are d-independent. Sampling is efficient if both µ and φn are tensor-product. Convergence results robust to noise > 0. No significant changes if y is vector-/function-valued A. Narayan (U. Utah) Emulators and surrogates
  • 40. Part II: Nonlinear approximation Sparse approximation Non-adapted basis functions, nonlinear approximation construction procedure A. Narayan (U. Utah) Emulators and surrogates
  • 41. Limited measurements u(x) + = y ≈ uN (x) = N n=1 cnφn(x), x ∈ D ⊆ Rd y ∈ R c ∈ RN V = span {φ1, . . . , φN } {(xm, ym)} M m=1 → {cn} N n=1 When d > 1, it is common for an a priori N = dim V to be very large. Least-squares: collecting M ∼ dim V measurements can be infeasible. A. Narayan (U. Utah) Emulators and surrogates
  • 42. Limited measurements u(x) + = y ≈ uN (x) = N n=1 cnφn(x), x ∈ D ⊆ Rd y ∈ R c ∈ RN V = span {φ1, . . . , φN } {(xm, ym)} M m=1 → {cn} N n=1 When d > 1, it is common for an a priori N = dim V to be very large. Least-squares: collecting M ∼ dim V measurements can be infeasible. What happens when M < N? The system Ac ≈ y, is now underdetermined. Unique solutions can be gauranteed if functional structure is imposed. A. Narayan (U. Utah) Emulators and surrogates
  • 43. Sparse approximation uN (x) = N n=1 cnφn(x) If M < N measurements are available, can we recover the largest M coefficients from the vector c? Assume y(x) = N n=1 cnφn(x) + (x), | | < η. The compressibility of y is measured by σV,s(c) = inf d∈RN d 0≤s c − d 1 , d 0 := {j ∈ {1, . . . , N} dj = 0 A. Narayan (U. Utah) Emulators and surrogates
  • 44. Sparse approximation y(x) = N n=1 cnφn(x) + (x), | | < η. ym = y(xm), A ≈ cy y is assumed to be compressible (i.e., c is assumed compressible) With a limited number, M, of measurements, seek to approximate the best s-term approximation of c. Ideally, s ∼ M. This is not possible if the sampling points are arbitrarily chosen. A. Narayan (U. Utah) Emulators and surrogates
  • 45. Compressed sensing It is possible to recover the best s-term approximation with high probability. Assume xm are iid sampled from µ, and that φn are L2 µ-orthonormal, and assume M CKµs log3 (s) log N, For any c ∈ RN , let ym = y(xm) = N n=1 cnφn(xm) + (xm), and assume | | ≤ η. Then, with probability exceeding 1 − N−γ log3 (s) , the solution c∗ to the optimization problem, min d 1 such that Ad − y 2 ≤ η √ M, satisfies c − c∗ 1 ≤ C1σV,s(c) + C2 √ s . Above, Kµ = max n=1,...,N φn L∞(D). [Rauhut 2010], [Rauhut, Ward 2010] A. Narayan (U. Utah) Emulators and surrogates
  • 46. Recovery of models with sparse representations Figure 2. Transition plots for uniform random variables for d = 2 (top row row). The left column corresponds to sampling from the random variable column the CSA method and the right column asymptotic sampling. For all low-dimensional and high-degree situations considered, CSA and performs significantly better than than probabilistic sampling acco optimization tolerances, and when th tolerance, the authors of [23] obtained We observe poor recovery since Kµ in sample requirement is poorly behaved: M CKµs log3 (s) log N, Kµ = max n=1,...,N φn L∞(D). This requirement is heavily dependent on µ. A. Narayan (U. Utah) Emulators and surrogates
  • 47. Better sampling Again, choosing a better sampling strategy ameliorates this issue. Sample xm ∼ µV , solve min d 1 such that Ad − y 2,w ≤ η √ M, where w are weights to make the discrete sampling unbiased. Exponential variables, Moreover the error in the approximation recovered by the asymptotic bounded sampling method for Beta variables increases with dimension. When d = 30 the asymptotic bounded sampling method fails to recover any polynomials regardless of the sparsity or the number of samples used. It is worth noting that case of Legendre polynomials sampled by Chebyshev distribution we have a complete independence of the order of approximation, which agrees with previous results in [42]. However there are numerical results in [23, 49] showing almost no recovery when using the Chebyshev sampling method in high-dimensions. With the help of the authors of [23] we have verified that the poor performance exhibited in the aforementioned papers is a result of numerical issues associated with the authors use of the `1 -minimization solver in SparseLab [16]. Specifically, the authors of [23] were using more lenient optimization tolerances, and when these tolerances were made tighter to match our optimization tolerance, the authors of [23] obtained results consistent with Figure 2. sampling method for Beta variables increases with dimension. When d = 30 the asymptotic bounded sampling method fails to recover any polynomials regardless of the sparsity or the number of samples used. It is worth noting that case of Legendre polynomials sampled by Chebyshev distribution we have a complete independence of the order of approximation, which agrees with previous results in [42]. However there are numerical results in [23, 49] showing almost no recovery when using the Chebyshev sampling method in high-dimensions. With the help of the authors of [23] we have verified that the poor performance exhibited in the aforementioned papers is a result of numerical issues associated with the authors use of the `1 -minimization solver in SparseLab [16]. Specifically, the authors of [23] were using more lenient optimization tolerances, and when these tolerances were made tighter to match our optimization tolerance, the authors of [23] obtained results consistent with Figure 2. A GENERALIZED SAMPLING AND PRECONDITIONING SCHEME F 101 102 103 10 14 10 13 10 12 10 11 10 10 10 9 10 8 10 7 10 6 10 5 10 4 10 3 10 2 10 1 Number of samples M `2error CSA MC 101 Figure 8. The e↵ect of dimension on the convergence of the CSA the di↵usion equation (28). (Left) 30th degree polynomial in 2 polynomial in 20 dimensions 10 4 10 3 10 2 10 1 rror CSA Asymptotic [Jakeman, AN 2017], [Guo, Zhou, Chen, AN 2016] A. Narayan (U. Utah) Emulators and surrogates
  • 48. Part III: Nonlinear approximation Dimension reduction/reduced modeling Adapted basis functions, nonlinear approximation construction procedure A. Narayan (U. Utah) Emulators and surrogates
  • 49. Dimension reduction u(x) + = y ≈ uN (x) = N n=1 cnφn(x), x ∈ D ⊆ Rd y ∈ RP cn ∈ RP A “sample” ym is a vector, possible of large size, P 1. In scientific models, P is also an indicator of the effort to obtain ym. Construct V and φ1, . . . , φN , by analyzing {(xm, ym)} M m=1 , (xm, ym) ∈ Rd × RP The φn are adapted to the data. Though φn has no explicit form, evaluating such functions can be much cheaper than gathering more data. A. Narayan (U. Utah) Emulators and surrogates
  • 50. Reduced basis methods Gather (xm, ym) from a scientific model. The reduced basis method (RBM) for nonlinear, adapted approximation, constructs the emulator uN (x) = N n=1 cnφn(x) = N n=1 yn n(x), Here: We need at least N = M data samples ym. n are cardinal Lagrange functions, satisfying n(xm) = δn,m. They have no explicit form. The n are defined implicitly from the scientific model. (Via a Galerkin procedure.) This is not POD. The space V = span{φn}N n=1 is constructed/defined from the data and the model. There is no reason to believe this is a good idea unless xm is chosen well! A. Narayan (U. Utah) Emulators and surrogates
  • 51. Reduced basis methods End goal: evaluation of surrogate uN should cost less than acquiring more data. Costs: Evaluting Lagrange functions n is the hard part – complexity usually scales like N3 . The full model ym is queried only at xm, and nowhere else. Details of computational efficiency of the surrogate uN depend on particular problem. In practice, N ∼ O(10). A. Narayan (U. Utah) Emulators and surrogates
  • 52. Lagrange functions -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 µ -3 -2 -1 0 1 2 The Lagrange functions uN 10,1 uN 10,2 uN 10,3 uN 10,4 uN 10,5 uN 10,6 uN 10,7 uN 10,8 uN 10,9 uN 10,10 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 µ 2 4 6 8 ˜∆10(µ) 5 10 15 20 25 30 35 40 Number of bases, N 10 -10 10 -5 10 0 -1 -0.5 0 0.5 1 µ 10 -6 10 -5 10 -4 10 -3 10 -2 10−4 × ˜∆10(µ) ||uN (µ) − uN 10,E3 (µ)||X S10 E3 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 µ -1 -0.5 0 0.5 1 1.5 The Lagrange functions uN 10,1 uN 10,2 uN 10,3 uN 10,4 uN 10,5 uN 10,6 uN 10,7 uN 10,8 uN 10,9 uN 10,10 3 ˜∆10(µ) A. Narayan (U. Utah) Emulators and surrogates
  • 53. RBM accuracy Does uN computed via RBM provide a good emulator for u? Depends on the sampling. Let u(xm) ∈ H. Suppose we choose xn+1 = arg max x∈D un(x) − u(x) H (This can be approximated without knowing u!) Then, u − uN L∞(D,H) σN (U), where U := u(x) x ∈ D ⊂ H, σN (U) := inf dim V =N sup v∈U inf vN ∈V v − vN H. [DeVore et al 2013], [Binev et al 2013] A. Narayan (U. Utah) Emulators and surrogates
  • 54. RBM accuracy Surrogates for nontrivial problems can be constructed. (−∆)s u(z; x) = f(z; ν), (z, x) ∈ Ω × D u(z; x) = 0, (z, x) ∈ ∂Ω × D Parameters/variables are x = (s, ν). 4 2 4 6 8 10 12 14 10 -10 10-5 10 0 tion UN associated to problem (42), where µ = (s, ⌫). 1 100 200 309 M 10 2 103 10 4 10 5 tMsolves s ∈ D1 M · tUN toffline + M · tU10 1 100 200 309 M 10 2 103 10 4 10 5 tMsolves s ∈ D2 M · tUN toffline + M · tU10 Figure 3: The cumulative computation time for M queries of the full order model uN and the RBM surrogate uN . On the left is for the case s 2 D1 with N = 7; on the right, s 2 D2 with [Antil, Chen, AN 2018] A. Narayan (U. Utah) Emulators and surrogates
  • 55. Building emulators Surrogate models can be enormously useful. Linear approximations with non-adapted basis functions “Easiest” to construct, with weakest accuracy guarantees. Querying surrogate generally very fast. Useful for analying large datasets Nonlinear approximations with non-adapted basis functions Harder to construct, but more general accuracy guarantees. Querying surrogate still very fast. Useful when data is limited. Nonlinear approximations with adapted basis functions Generally very hard to construct. Very attractive accuracy bounds, when possible to certify Depend heavily on data, model, and the transparency of the model. A. Narayan (U. Utah) Emulators and surrogates
  • 56. Building emulators Surrogate models can be enormously useful. Linear approximations with non-adapted basis functions “Easiest” to construct, with weakest accuracy guarantees. Querying surrogate generally very fast. Useful for analying large datasets Nonlinear approximations with non-adapted basis functions Harder to construct, but more general accuracy guarantees. Querying surrogate still very fast. Useful when data is limited. Nonlinear approximations with adapted basis functions Generally very hard to construct. Very attractive accuracy bounds, when possible to certify Depend heavily on data, model, and the transparency of the model. Challenges: high dimensionality (d, P, or N) adaptivity and hierarchical constructions A. Narayan (U. Utah) Emulators and surrogates
  • 57. mathematics of reduced order models algorithms for approximation and complexity reduction computational statistics and data-driven techniques https://icerm.brown.edu/programs/sp-s20/ A. Narayan (U. Utah) Emulators and surrogates
  • 58. References Chkifa, Cohen, Migliorati, Nobile, Tempone, ”Discrete least squares polynomial approximation with random evaluations – application to parametric and stochastic elliptic PDEs”, ESAIM: Mathematical Modelling and Numerical Analysis, 49:3 (2015) Cohen, Davenport, & Leviatan, ”On the Stability and Accuracy of Least Squares Approximations”, Foundations of Computational Mathematics, 13:5 (2013) Cohen & Migliorati, ”Optimal weighted least-squares methods”, arXiv:1608.00512 [math, stat] Jakeman, Narayan, & Zhou, ”A Christoffel function weighted least squares algorithm for collocation approximations”, Mathematics of Computation, 86:306 (2017) Narayan, ”Computation of Induced Orthogonal Polynomial Distributions”, arXiv:1704.08465 [math] (2017) Narayan & Zhou, ”Stochastic Collocation on Unstructured Multivariate Meshes”, Communications in Computational Physics, 18:1 (2015) A. Narayan (U. Utah) Emulators and surrogates