SlideShare a Scribd company logo
1 of 16
Download to read offline
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Boosted Tree-based Multinomial Logit Model for
Aggregated Market Data
Jianqiang (Jay) Wang & Trevor Hastie
Hewlett-Packard Labs & Stanford University
Dec 2, 2012
Disclaimer: I, myself, take sole responsibility for any errors and omissions in this presentation.
1 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Hewlett-Packard Labs
HPL Charter:
DELIVER; CREATE; ADVANCE; ENGAGE
Information Analytics Lab:
2 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Statistical Demand Modeling
3 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Pricing and Portfolio Management
Predictive analytics-based PPM decision support system.
2012 INFORMS Revenue Management & Pricing Practice Award.
Demand
How do consumers value products?
Product Selection and Pricing
What products should we offer? What is the right pricing?
Competitive Product Similarity
What products are we competing with on the market?
Leveraging Intelligence
Can we infer market intelligence from current prices, and
learn?
4 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Estimating Aggregated Market Demand
Aggregated mobile computer sales data on all brands.
Market sales data reveals customer selection.
Aggregated mobile PC sales.
Brands, country, region, attributes, period, channel, price, volume.
Complexity of model estimation:
40+ different key features (memory, CPU, display, storage, OS, ...).
Price sensitivity varies with attributes, time, and region.
High-dimensional prediction problem.
5 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Discrete Choice Model
Modeling Sales Volume vs Consumer choice (McFadden 1974):
Choice set: products to choose from.
Utility : overall attractiveness given attributes, brand and price.
Better attributes, higher utility; higher price, lower utility.
Challenges:
Sparse selection.
Nonlinearity.
Interactions among (attributes, price).
Semiparametric Multinomial Logit Model (MNL):
Linear MNLs: Train (2003); Semiparametric MNLs: p-splines (Tutz & Scholz 2004).
Flexibly model customers’ valuation without specifying a functional form.
Estimation: Functional gradient boosting with partitioned regression trees as base learners.
6 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Aggregated Market Multinomial Logit Model
Single market with K products; products i = 1, · · · , K with sales volumn
(n1, · · · , nM ); latent utilities
ui = fi + i .
Assuming i
iid
∼ standard Gumbel distn, utility maximization leads to
pi =
exp(fi )
K
i=1
exp(fi )
.
Minimize −2 log (multinomial likelihood):
φ(f) = −2
K
i=1
ni log(g(fi )) + 2N log



K
i=1
g(fi )



+ const.
g(·) link function, e.g., g(u) = exp(u).
7 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Model Variations
Notation: si – attributes, brand and channel; xi = (1, xi ) , xi – price.
Utility Specifications:
Varying coefficient-MNL (price*attribute interaction):
fi = xi β(si ).
Partially linear-MNL (price & attribute additive):
fi = β0(si ) + xi β1.
Nonparametric-MNL:
fi = β(si , xi ).
Boosted trees:
Partition the products into homogeneous groups in a way that respects the mean utility function..
Iteratively fits simple trees to explain errors not captured in the previous iteration.
8 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Building Block: VC Trees
Underlying VCM model:
ξi = xi β(si ) + i ,
Piecewise constant approximation:
ξi =
M
m=1
xi βmI(si ∈Cm) + i ,
M: number of partitions.
{Cm}M
m=1: a partition of the space of si .
Piecewise constant approximation to the unknown high-dimensional function &
data-driven partitioning method to obtain homogeneous regression relationships.
Algorithm:
Heuristics: greedy algorithm based on binary splits of the space of si (similar to CART).
Splitting criterion: reduction in SSE.
9 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Boosted VC-MNL
Boosted VC-MNL: φ(f) = −2 K
i=1 ni log(g(xi β(si ))) + 2N log K
i=1 g(xi β(si )) + const.
1 Start with naive fit ˆf
(0)
= (x1β
(0)
, · · · , xK β
(0)
) .
2 For b = 1, · · · , B, repeat:
Compute the “pseudo observations”: ξi = − ∂φ
∂fi f =ˆf (b−1)
.
Fit ξi on si and xi using the “PartReg” algorithm to obtain partitions (C
(b)
1 , · · · , C
(b)
M
).
Let zi = (I
(si ∈C
(b)
1
)
, · · · , I
(si ∈C
(b)
M
)
, xi I
(si ∈C
(b)
1
)
, · · · , xi I
(si ∈C
(b)
M
)
) , and use IRLS to
estimate β
(b)
by minimizing
J(β
(b)
) = −2
K
i=1
ni log(g(ˆf
(b−1)
i
+ zi β
(b)
)) + 2N log



K
i=1
g(ˆf
(b−1)
i
+ zi β
(b)
)



.
Update the fitted model by ˆf (b)
= ˆf (b−1)
+ ν M
m=1 β
(b)
0m + β
(b)
1mxi I
(si ∈C
(b)
m )
.
3 Output the fitted model ˆf = ˆf (B).
10 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Boosted VC-MNL
Start with naive fit: e.g., simple linear MNL.
Begin the iteration process:
Compute pseudo observations/residuals.
Fit an appropriate tree to predict pseudo residualts.
Generate design matrix based on tree partitions, and fit linear MNL model.
Addtive model of trees, not of predictors.
Iteratively fit linear MNL models based on data-driven piecewise constant
“bases”.
11 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Mobile Computer Sales in Australia
6 months, 5 states; 30 choice sets (25 training, 5 test); use price residuals
instead of price.
Varying coefficient-MNL:
fi = xi β(si ).
Partially linear-MNL:
fi = β0(si ) + xi β1.
Nonparametric-MNL:
fi = β(si , xi ).
0 200 400 600 800 1000
0.00.20.40.60.81.0
Varying coefficient−MNL, Boosted
Iterations
R2
Training
Test
0 200 400 600 800 1000
0.00.20.40.60.81.0
Partially linear, Boosted
Iterations
R2
Training
Test
0 200 400 600 800 1000
0.00.20.40.60.81.0
Nonparametric, Boosted
Iterations
R2
Training
Test
12 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Competitor Method – Elastic Net MNL
Models: fi = xi β(si ).
Linear-MNL: linear β(si ).
Quadratic-MNL (first-order interaction).
Quadratic-MNL: Initial features si .
⇒ Quadratic & first-order interaction among si , obtain design matrix zi .
⇒ Linear specification: β0(si ) = zi γ0 and β1(si ) = zi γ1.
Elastic net (Zou & Hastie 2005) MNL:
arg min
γ0,γ1
−2
K
i=1
ni log(g(zi γ0 + (zi xi )γ1)) + 2N log



K
i=1
g(zi γ0 + (zi xi )γ1)



+λ



α
i,j
|γij | +
(1 − α)
2 i,j
γ
2
ij



α = 0: Ridge regression; α = 1: LASSO.
g(·) : link function.
Sparse and stable coefficient estimates, penalized IRLS.
13 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Summary of Results
Utility Optimal R2
Interactions
Specification
Estimation
Training Test
Time (min)
Among attributes
(α = 1) 399 .357 .17 X
Linear
(α = 1
2
) .419 .379 .48 X
(α = 1)
penalized IRLS
.582 .499 76.91 1st
-order
Quadratic
(α = 1
2
) .554 .53 52.78 1st
-order
Varying-coef. .734 .697 186.47 (B=1000)
Partially linear boosted trees .493 .455 24.63 (B=1000) 2nd
-order (M=4)
Nonparametric .52 .502 23.43 (B=1000)
M – size of each base tree; B– the number of boosting iterations
Nonparametric MNL specifies a larger model space than VC-MNL, but piecewise constant trees fails to find the
particular interactions.
14 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Discussion
Semiparametric MNL models, estimated by boosted tree methods.
Learning from large-scale market data to a) make predictions and b) gain
insights: econometrics & statistical learning.
Statistical questions:
Assessing errors in R2
and coefficient surface.
Split selection in tree partitioning (variable importance).
Model validation & diagnostics (standardized pseudo residuals).
Choice of link functions.
15 / 16
Motivation Aggregated Market Multinomial Logit Model Application to Australian Data
Jianqiang (Jay) Wang
Information Analytics Lab
Hewlett-Packard Labs
jianqiang.jay.wang@hp.com
Thank you very much!
16 / 16

More Related Content

What's hot

Simple representations for learning: factorizations and similarities
Simple representations for learning: factorizations and similarities Simple representations for learning: factorizations and similarities
Simple representations for learning: factorizations and similarities
Gael Varoquaux
 
Sparse Kernel Learning for Image Annotation
Sparse Kernel Learning for Image AnnotationSparse Kernel Learning for Image Annotation
Sparse Kernel Learning for Image Annotation
Sean Moran
 
All Pair Shortest Path Algorithm – Parallel Implementation and Analysis
All Pair Shortest Path Algorithm – Parallel Implementation and AnalysisAll Pair Shortest Path Algorithm – Parallel Implementation and Analysis
All Pair Shortest Path Algorithm – Parallel Implementation and Analysis
Inderjeet Singh
 

What's hot (20)

Simple representations for learning: factorizations and similarities
Simple representations for learning: factorizations and similarities Simple representations for learning: factorizations and similarities
Simple representations for learning: factorizations and similarities
 
VAE-type Deep Generative Models
VAE-type Deep Generative ModelsVAE-type Deep Generative Models
VAE-type Deep Generative Models
 
Predicting organic reaction outcomes with weisfeiler lehman network
Predicting organic reaction outcomes with weisfeiler lehman networkPredicting organic reaction outcomes with weisfeiler lehman network
Predicting organic reaction outcomes with weisfeiler lehman network
 
MUMS: Bayesian, Fiducial, and Frequentist Conference - Model Selection in the...
MUMS: Bayesian, Fiducial, and Frequentist Conference - Model Selection in the...MUMS: Bayesian, Fiducial, and Frequentist Conference - Model Selection in the...
MUMS: Bayesian, Fiducial, and Frequentist Conference - Model Selection in the...
 
XGBoost & LightGBM
XGBoost & LightGBMXGBoost & LightGBM
XGBoost & LightGBM
 
NIPS2017 Few-shot Learning and Graph Convolution
NIPS2017 Few-shot Learning and Graph ConvolutionNIPS2017 Few-shot Learning and Graph Convolution
NIPS2017 Few-shot Learning and Graph Convolution
 
Support Vector Machines (SVM)
Support Vector Machines (SVM)Support Vector Machines (SVM)
Support Vector Machines (SVM)
 
Sparse Kernel Learning for Image Annotation
Sparse Kernel Learning for Image AnnotationSparse Kernel Learning for Image Annotation
Sparse Kernel Learning for Image Annotation
 
Gradient Estimation Using Stochastic Computation Graphs
Gradient Estimation Using Stochastic Computation GraphsGradient Estimation Using Stochastic Computation Graphs
Gradient Estimation Using Stochastic Computation Graphs
 
All Pair Shortest Path Algorithm – Parallel Implementation and Analysis
All Pair Shortest Path Algorithm – Parallel Implementation and AnalysisAll Pair Shortest Path Algorithm – Parallel Implementation and Analysis
All Pair Shortest Path Algorithm – Parallel Implementation and Analysis
 
A Hough Transform Based On a Map-Reduce Algorithm
A Hough Transform Based On a Map-Reduce AlgorithmA Hough Transform Based On a Map-Reduce Algorithm
A Hough Transform Based On a Map-Reduce Algorithm
 
Iclr2016 vaeまとめ
Iclr2016 vaeまとめIclr2016 vaeまとめ
Iclr2016 vaeまとめ
 
Gradient boosting in practice: a deep dive into xgboost
Gradient boosting in practice: a deep dive into xgboostGradient boosting in practice: a deep dive into xgboost
Gradient boosting in practice: a deep dive into xgboost
 
Scaling out logistic regression with Spark
Scaling out logistic regression with SparkScaling out logistic regression with Spark
Scaling out logistic regression with Spark
 
Accelerating Random Forests in Scikit-Learn
Accelerating Random Forests in Scikit-LearnAccelerating Random Forests in Scikit-Learn
Accelerating Random Forests in Scikit-Learn
 
Higher-order Factorization Machines(第5回ステアラボ人工知能セミナー)
Higher-order Factorization Machines(第5回ステアラボ人工知能セミナー)Higher-order Factorization Machines(第5回ステアラボ人工知能セミナー)
Higher-order Factorization Machines(第5回ステアラボ人工知能セミナー)
 
Bayesian Dark Knowledge and Matrix Factorization
Bayesian Dark Knowledge and Matrix FactorizationBayesian Dark Knowledge and Matrix Factorization
Bayesian Dark Knowledge and Matrix Factorization
 
Goodfellow, Bengio, Couville (2016) "Deep Learning", Chap. 7
Goodfellow, Bengio, Couville (2016) "Deep Learning", Chap. 7Goodfellow, Bengio, Couville (2016) "Deep Learning", Chap. 7
Goodfellow, Bengio, Couville (2016) "Deep Learning", Chap. 7
 
Traffic flow modeling on road networks using Hamilton-Jacobi equations
Traffic flow modeling on road networks using Hamilton-Jacobi equationsTraffic flow modeling on road networks using Hamilton-Jacobi equations
Traffic flow modeling on road networks using Hamilton-Jacobi equations
 
그림 그리는 AI
그림 그리는 AI그림 그리는 AI
그림 그리는 AI
 

Viewers also liked

1.5.1 measures basic concepts
1.5.1 measures basic concepts1.5.1 measures basic concepts
1.5.1 measures basic concepts
A M
 
(마더세이프 라운드) Logistic regression
(마더세이프 라운드) Logistic regression(마더세이프 라운드) Logistic regression
(마더세이프 라운드) Logistic regression
mothersafe
 

Viewers also liked (20)

1.5.1 measures basic concepts
1.5.1 measures basic concepts1.5.1 measures basic concepts
1.5.1 measures basic concepts
 
Logistic regression
Logistic regressionLogistic regression
Logistic regression
 
Logistic regression with SPSS examples
Logistic regression with SPSS examplesLogistic regression with SPSS examples
Logistic regression with SPSS examples
 
Logistic Regression/Markov Chain presentation
Logistic Regression/Markov Chain presentationLogistic Regression/Markov Chain presentation
Logistic Regression/Markov Chain presentation
 
Ordinal Logistic Regression
Ordinal Logistic RegressionOrdinal Logistic Regression
Ordinal Logistic Regression
 
Transparency7
Transparency7Transparency7
Transparency7
 
(마더세이프 라운드) Logistic regression
(마더세이프 라운드) Logistic regression(마더세이프 라운드) Logistic regression
(마더세이프 라운드) Logistic regression
 
Generalized Logistic Regression - by example (Anthony Kilili)
Generalized Logistic Regression - by example (Anthony Kilili)Generalized Logistic Regression - by example (Anthony Kilili)
Generalized Logistic Regression - by example (Anthony Kilili)
 
Mode Choice analysis for work trips using Multinomial Logit model for Windsor...
Mode Choice analysis for work trips using Multinomial Logit model for Windsor...Mode Choice analysis for work trips using Multinomial Logit model for Windsor...
Mode Choice analysis for work trips using Multinomial Logit model for Windsor...
 
Intro to Logistic Regression
Intro to Logistic RegressionIntro to Logistic Regression
Intro to Logistic Regression
 
Logistic Regression: Behind the Scenes
Logistic Regression: Behind the ScenesLogistic Regression: Behind the Scenes
Logistic Regression: Behind the Scenes
 
From logistic regression to linear chain CRF
From logistic regression to linear chain CRFFrom logistic regression to linear chain CRF
From logistic regression to linear chain CRF
 
Choice Models
Choice ModelsChoice Models
Choice Models
 
4.5. logistic regression
4.5. logistic regression4.5. logistic regression
4.5. logistic regression
 
Binary Logistic Regression Example
Binary Logistic Regression ExampleBinary Logistic Regression Example
Binary Logistic Regression Example
 
ESL 4.4.3-4.5: Logistic Reression (contd.) and Separating Hyperplane
ESL 4.4.3-4.5: Logistic Reression (contd.) and Separating HyperplaneESL 4.4.3-4.5: Logistic Reression (contd.) and Separating Hyperplane
ESL 4.4.3-4.5: Logistic Reression (contd.) and Separating Hyperplane
 
Logistic regression for ordered dependant variable with more than 2 levels
Logistic regression for ordered dependant variable with more than 2 levelsLogistic regression for ordered dependant variable with more than 2 levels
Logistic regression for ordered dependant variable with more than 2 levels
 
Logistic regression (blyth 2006) (simplified)
Logistic regression (blyth 2006) (simplified)Logistic regression (blyth 2006) (simplified)
Logistic regression (blyth 2006) (simplified)
 
Logistic regression
Logistic regressionLogistic regression
Logistic regression
 
Probit and logit model
Probit and logit modelProbit and logit model
Probit and logit model
 

Similar to Boosted Tree-based Multinomial Logit Model for Aggregated Market Data

AggNet: Deep Learning from Crowds
AggNet: Deep Learning from CrowdsAggNet: Deep Learning from Crowds
AggNet: Deep Learning from Crowds
Shadi Nabil Albarqouni
 
Iwsm2014 an analogy-based approach to estimation of software development ef...
Iwsm2014   an analogy-based approach to estimation of software development ef...Iwsm2014   an analogy-based approach to estimation of software development ef...
Iwsm2014 an analogy-based approach to estimation of software development ef...
Nesma
 

Similar to Boosted Tree-based Multinomial Logit Model for Aggregated Market Data (20)

GDRR Opening Workshop - Modeling Approaches for High-Frequency Financial Time...
GDRR Opening Workshop - Modeling Approaches for High-Frequency Financial Time...GDRR Opening Workshop - Modeling Approaches for High-Frequency Financial Time...
GDRR Opening Workshop - Modeling Approaches for High-Frequency Financial Time...
 
Asynchronous Stochastic Optimization, New Analysis and Algorithms
Asynchronous Stochastic Optimization, New Analysis and AlgorithmsAsynchronous Stochastic Optimization, New Analysis and Algorithms
Asynchronous Stochastic Optimization, New Analysis and Algorithms
 
Graphical Model Selection for Big Data
Graphical Model Selection for Big DataGraphical Model Selection for Big Data
Graphical Model Selection for Big Data
 
Applying Linear Optimization Using GLPK
Applying Linear Optimization Using GLPKApplying Linear Optimization Using GLPK
Applying Linear Optimization Using GLPK
 
Keefe Murphy - IMIFA - CASI 2017
Keefe Murphy - IMIFA - CASI 2017Keefe Murphy - IMIFA - CASI 2017
Keefe Murphy - IMIFA - CASI 2017
 
sheet6.pdf
sheet6.pdfsheet6.pdf
sheet6.pdf
 
doc6.pdf
doc6.pdfdoc6.pdf
doc6.pdf
 
paper6.pdf
paper6.pdfpaper6.pdf
paper6.pdf
 
lecture5.pdf
lecture5.pdflecture5.pdf
lecture5.pdf
 
AggNet: Deep Learning from Crowds
AggNet: Deep Learning from CrowdsAggNet: Deep Learning from Crowds
AggNet: Deep Learning from Crowds
 
Self-sampling Strategies for Multimemetic Algorithms in Unstable Computationa...
Self-sampling Strategies for Multimemetic Algorithms in Unstable Computationa...Self-sampling Strategies for Multimemetic Algorithms in Unstable Computationa...
Self-sampling Strategies for Multimemetic Algorithms in Unstable Computationa...
 
CLIM Program: Remote Sensing Workshop, Optimization for Distributed Data Syst...
CLIM Program: Remote Sensing Workshop, Optimization for Distributed Data Syst...CLIM Program: Remote Sensing Workshop, Optimization for Distributed Data Syst...
CLIM Program: Remote Sensing Workshop, Optimization for Distributed Data Syst...
 
Bat algorithm for Topology Optimization in Microelectronic Applications
Bat algorithm for Topology Optimization in Microelectronic ApplicationsBat algorithm for Topology Optimization in Microelectronic Applications
Bat algorithm for Topology Optimization in Microelectronic Applications
 
Count Data Models in SAS
Count Data Models in SASCount Data Models in SAS
Count Data Models in SAS
 
Presentation eng
Presentation engPresentation eng
Presentation eng
 
Predicting Short Term Movements of Stock Prices: A Two-Stage L1-Penalized Model
Predicting Short Term Movements of Stock Prices: A Two-Stage L1-Penalized ModelPredicting Short Term Movements of Stock Prices: A Two-Stage L1-Penalized Model
Predicting Short Term Movements of Stock Prices: A Two-Stage L1-Penalized Model
 
Iwsm2014 an analogy-based approach to estimation of software development ef...
Iwsm2014   an analogy-based approach to estimation of software development ef...Iwsm2014   an analogy-based approach to estimation of software development ef...
Iwsm2014 an analogy-based approach to estimation of software development ef...
 
SAT based planning for multiagent systems
SAT based planning for multiagent systemsSAT based planning for multiagent systems
SAT based planning for multiagent systems
 
Reading revue of "Inferring Multiple Graphical Structures"
Reading revue of "Inferring Multiple Graphical Structures"Reading revue of "Inferring Multiple Graphical Structures"
Reading revue of "Inferring Multiple Graphical Structures"
 
2019 Fall Series: Postdoc Seminars - Special Guest Lecture, Attacking the Cur...
2019 Fall Series: Postdoc Seminars - Special Guest Lecture, Attacking the Cur...2019 Fall Series: Postdoc Seminars - Special Guest Lecture, Attacking the Cur...
2019 Fall Series: Postdoc Seminars - Special Guest Lecture, Attacking the Cur...
 

More from Jay (Jianqiang) Wang

More from Jay (Jianqiang) Wang (10)

The Practice of Data Driven Products in Kuaishou
The Practice of Data Driven Products in KuaishouThe Practice of Data Driven Products in Kuaishou
The Practice of Data Driven Products in Kuaishou
 
Artificial Intelligence in fashion -- Combining Statistics and Expert Human J...
Artificial Intelligence in fashion -- Combining Statistics and Expert Human J...Artificial Intelligence in fashion -- Combining Statistics and Expert Human J...
Artificial Intelligence in fashion -- Combining Statistics and Expert Human J...
 
Making data-informed decisions and building intelligent products (Chinese)
Making data-informed decisions and building intelligent products (Chinese)Making data-informed decisions and building intelligent products (Chinese)
Making data-informed decisions and building intelligent products (Chinese)
 
Notes on Machine Learning and Data-centric Startups
Notes on Machine Learning and Data-centric StartupsNotes on Machine Learning and Data-centric Startups
Notes on Machine Learning and Data-centric Startups
 
Introduction to data science and its application in online advertising
Introduction to data science and its application in online advertisingIntroduction to data science and its application in online advertising
Introduction to data science and its application in online advertising
 
How to prepare for data science interviews
How to prepare for data science interviewsHow to prepare for data science interviews
How to prepare for data science interviews
 
Introduction to data science and candidate data science projects
Introduction to data science and candidate data science projectsIntroduction to data science and candidate data science projects
Introduction to data science and candidate data science projects
 
Multivariate outlier detection
Multivariate outlier detectionMultivariate outlier detection
Multivariate outlier detection
 
Multivariate outlier detection
Multivariate outlier detectionMultivariate outlier detection
Multivariate outlier detection
 
A Bayesian Approach to Estimating Agricultual Yield Based on Multiple Repeat...
A Bayesian Approach to Estimating Agricultual Yield Based on Multiple Repeat...A Bayesian Approach to Estimating Agricultual Yield Based on Multiple Repeat...
A Bayesian Approach to Estimating Agricultual Yield Based on Multiple Repeat...
 

Recently uploaded

Cytotec in Jeddah+966572737505) get unwanted pregnancy kit Riyadh
Cytotec in Jeddah+966572737505) get unwanted pregnancy kit RiyadhCytotec in Jeddah+966572737505) get unwanted pregnancy kit Riyadh
Cytotec in Jeddah+966572737505) get unwanted pregnancy kit Riyadh
Abortion pills in Riyadh +966572737505 get cytotec
 
Abortion pills in Doha Qatar (+966572737505 ! Get Cytotec
Abortion pills in Doha Qatar (+966572737505 ! Get CytotecAbortion pills in Doha Qatar (+966572737505 ! Get Cytotec
Abortion pills in Doha Qatar (+966572737505 ! Get Cytotec
Abortion pills in Riyadh +966572737505 get cytotec
 
Jual Obat Aborsi Surabaya ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
Jual Obat Aborsi Surabaya ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...Jual Obat Aborsi Surabaya ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
Jual Obat Aborsi Surabaya ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
ZurliaSoop
 
Jual Cytotec Asli Obat Aborsi No. 1 Paling Manjur
Jual Cytotec Asli Obat Aborsi No. 1 Paling ManjurJual Cytotec Asli Obat Aborsi No. 1 Paling Manjur
Jual Cytotec Asli Obat Aborsi No. 1 Paling Manjur
ptikerjasaptiker
 
Top profile Call Girls In Purnia [ 7014168258 ] Call Me For Genuine Models We...
Top profile Call Girls In Purnia [ 7014168258 ] Call Me For Genuine Models We...Top profile Call Girls In Purnia [ 7014168258 ] Call Me For Genuine Models We...
Top profile Call Girls In Purnia [ 7014168258 ] Call Me For Genuine Models We...
nirzagarg
 
Reconciling Conflicting Data Curation Actions: Transparency Through Argument...
Reconciling Conflicting Data Curation Actions:  Transparency Through Argument...Reconciling Conflicting Data Curation Actions:  Transparency Through Argument...
Reconciling Conflicting Data Curation Actions: Transparency Through Argument...
Bertram Ludäscher
 
Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...
Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...
Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...
gajnagarg
 
Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...
Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...
Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...
Klinik kandungan
 
怎样办理圣路易斯大学毕业证(SLU毕业证书)成绩单学校原版复制
怎样办理圣路易斯大学毕业证(SLU毕业证书)成绩单学校原版复制怎样办理圣路易斯大学毕业证(SLU毕业证书)成绩单学校原版复制
怎样办理圣路易斯大学毕业证(SLU毕业证书)成绩单学校原版复制
vexqp
 
Lecture_2_Deep_Learning_Overview-newone1
Lecture_2_Deep_Learning_Overview-newone1Lecture_2_Deep_Learning_Overview-newone1
Lecture_2_Deep_Learning_Overview-newone1
ranjankumarbehera14
 
PLE-statistics document for primary schs
PLE-statistics document for primary schsPLE-statistics document for primary schs
PLE-statistics document for primary schs
cnajjemba
 
怎样办理圣地亚哥州立大学毕业证(SDSU毕业证书)成绩单学校原版复制
怎样办理圣地亚哥州立大学毕业证(SDSU毕业证书)成绩单学校原版复制怎样办理圣地亚哥州立大学毕业证(SDSU毕业证书)成绩单学校原版复制
怎样办理圣地亚哥州立大学毕业证(SDSU毕业证书)成绩单学校原版复制
vexqp
 
Gartner's Data Analytics Maturity Model.pptx
Gartner's Data Analytics Maturity Model.pptxGartner's Data Analytics Maturity Model.pptx
Gartner's Data Analytics Maturity Model.pptx
chadhar227
 

Recently uploaded (20)

Cytotec in Jeddah+966572737505) get unwanted pregnancy kit Riyadh
Cytotec in Jeddah+966572737505) get unwanted pregnancy kit RiyadhCytotec in Jeddah+966572737505) get unwanted pregnancy kit Riyadh
Cytotec in Jeddah+966572737505) get unwanted pregnancy kit Riyadh
 
Switzerland Constitution 2002.pdf.........
Switzerland Constitution 2002.pdf.........Switzerland Constitution 2002.pdf.........
Switzerland Constitution 2002.pdf.........
 
Abortion pills in Doha Qatar (+966572737505 ! Get Cytotec
Abortion pills in Doha Qatar (+966572737505 ! Get CytotecAbortion pills in Doha Qatar (+966572737505 ! Get Cytotec
Abortion pills in Doha Qatar (+966572737505 ! Get Cytotec
 
Jual Obat Aborsi Surabaya ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
Jual Obat Aborsi Surabaya ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...Jual Obat Aborsi Surabaya ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
Jual Obat Aborsi Surabaya ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
 
Jual Cytotec Asli Obat Aborsi No. 1 Paling Manjur
Jual Cytotec Asli Obat Aborsi No. 1 Paling ManjurJual Cytotec Asli Obat Aborsi No. 1 Paling Manjur
Jual Cytotec Asli Obat Aborsi No. 1 Paling Manjur
 
Sequential and reinforcement learning for demand side management by Margaux B...
Sequential and reinforcement learning for demand side management by Margaux B...Sequential and reinforcement learning for demand side management by Margaux B...
Sequential and reinforcement learning for demand side management by Margaux B...
 
Top profile Call Girls In Purnia [ 7014168258 ] Call Me For Genuine Models We...
Top profile Call Girls In Purnia [ 7014168258 ] Call Me For Genuine Models We...Top profile Call Girls In Purnia [ 7014168258 ] Call Me For Genuine Models We...
Top profile Call Girls In Purnia [ 7014168258 ] Call Me For Genuine Models We...
 
Reconciling Conflicting Data Curation Actions: Transparency Through Argument...
Reconciling Conflicting Data Curation Actions:  Transparency Through Argument...Reconciling Conflicting Data Curation Actions:  Transparency Through Argument...
Reconciling Conflicting Data Curation Actions: Transparency Through Argument...
 
7. Epi of Chronic respiratory diseases.ppt
7. Epi of Chronic respiratory diseases.ppt7. Epi of Chronic respiratory diseases.ppt
7. Epi of Chronic respiratory diseases.ppt
 
Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...
Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...
Top profile Call Girls In dimapur [ 7014168258 ] Call Me For Genuine Models W...
 
Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...
Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...
Jual obat aborsi Bandung ( 085657271886 ) Cytote pil telat bulan penggugur ka...
 
怎样办理圣路易斯大学毕业证(SLU毕业证书)成绩单学校原版复制
怎样办理圣路易斯大学毕业证(SLU毕业证书)成绩单学校原版复制怎样办理圣路易斯大学毕业证(SLU毕业证书)成绩单学校原版复制
怎样办理圣路易斯大学毕业证(SLU毕业证书)成绩单学校原版复制
 
Lecture_2_Deep_Learning_Overview-newone1
Lecture_2_Deep_Learning_Overview-newone1Lecture_2_Deep_Learning_Overview-newone1
Lecture_2_Deep_Learning_Overview-newone1
 
PLE-statistics document for primary schs
PLE-statistics document for primary schsPLE-statistics document for primary schs
PLE-statistics document for primary schs
 
怎样办理圣地亚哥州立大学毕业证(SDSU毕业证书)成绩单学校原版复制
怎样办理圣地亚哥州立大学毕业证(SDSU毕业证书)成绩单学校原版复制怎样办理圣地亚哥州立大学毕业证(SDSU毕业证书)成绩单学校原版复制
怎样办理圣地亚哥州立大学毕业证(SDSU毕业证书)成绩单学校原版复制
 
Dubai Call Girls Peeing O525547819 Call Girls Dubai
Dubai Call Girls Peeing O525547819 Call Girls DubaiDubai Call Girls Peeing O525547819 Call Girls Dubai
Dubai Call Girls Peeing O525547819 Call Girls Dubai
 
DATA SUMMIT 24 Building Real-Time Pipelines With FLaNK
DATA SUMMIT 24  Building Real-Time Pipelines With FLaNKDATA SUMMIT 24  Building Real-Time Pipelines With FLaNK
DATA SUMMIT 24 Building Real-Time Pipelines With FLaNK
 
Harnessing the Power of GenAI for BI and Reporting.pptx
Harnessing the Power of GenAI for BI and Reporting.pptxHarnessing the Power of GenAI for BI and Reporting.pptx
Harnessing the Power of GenAI for BI and Reporting.pptx
 
Gartner's Data Analytics Maturity Model.pptx
Gartner's Data Analytics Maturity Model.pptxGartner's Data Analytics Maturity Model.pptx
Gartner's Data Analytics Maturity Model.pptx
 
Vadodara 💋 Call Girl 7737669865 Call Girls in Vadodara Escort service book now
Vadodara 💋 Call Girl 7737669865 Call Girls in Vadodara Escort service book nowVadodara 💋 Call Girl 7737669865 Call Girls in Vadodara Escort service book now
Vadodara 💋 Call Girl 7737669865 Call Girls in Vadodara Escort service book now
 

Boosted Tree-based Multinomial Logit Model for Aggregated Market Data

  • 1. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Boosted Tree-based Multinomial Logit Model for Aggregated Market Data Jianqiang (Jay) Wang & Trevor Hastie Hewlett-Packard Labs & Stanford University Dec 2, 2012 Disclaimer: I, myself, take sole responsibility for any errors and omissions in this presentation. 1 / 16
  • 2. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Hewlett-Packard Labs HPL Charter: DELIVER; CREATE; ADVANCE; ENGAGE Information Analytics Lab: 2 / 16
  • 3. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Statistical Demand Modeling 3 / 16
  • 4. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Pricing and Portfolio Management Predictive analytics-based PPM decision support system. 2012 INFORMS Revenue Management & Pricing Practice Award. Demand How do consumers value products? Product Selection and Pricing What products should we offer? What is the right pricing? Competitive Product Similarity What products are we competing with on the market? Leveraging Intelligence Can we infer market intelligence from current prices, and learn? 4 / 16
  • 5. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Estimating Aggregated Market Demand Aggregated mobile computer sales data on all brands. Market sales data reveals customer selection. Aggregated mobile PC sales. Brands, country, region, attributes, period, channel, price, volume. Complexity of model estimation: 40+ different key features (memory, CPU, display, storage, OS, ...). Price sensitivity varies with attributes, time, and region. High-dimensional prediction problem. 5 / 16
  • 6. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Discrete Choice Model Modeling Sales Volume vs Consumer choice (McFadden 1974): Choice set: products to choose from. Utility : overall attractiveness given attributes, brand and price. Better attributes, higher utility; higher price, lower utility. Challenges: Sparse selection. Nonlinearity. Interactions among (attributes, price). Semiparametric Multinomial Logit Model (MNL): Linear MNLs: Train (2003); Semiparametric MNLs: p-splines (Tutz & Scholz 2004). Flexibly model customers’ valuation without specifying a functional form. Estimation: Functional gradient boosting with partitioned regression trees as base learners. 6 / 16
  • 7. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Aggregated Market Multinomial Logit Model Single market with K products; products i = 1, · · · , K with sales volumn (n1, · · · , nM ); latent utilities ui = fi + i . Assuming i iid ∼ standard Gumbel distn, utility maximization leads to pi = exp(fi ) K i=1 exp(fi ) . Minimize −2 log (multinomial likelihood): φ(f) = −2 K i=1 ni log(g(fi )) + 2N log    K i=1 g(fi )    + const. g(·) link function, e.g., g(u) = exp(u). 7 / 16
  • 8. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Model Variations Notation: si – attributes, brand and channel; xi = (1, xi ) , xi – price. Utility Specifications: Varying coefficient-MNL (price*attribute interaction): fi = xi β(si ). Partially linear-MNL (price & attribute additive): fi = β0(si ) + xi β1. Nonparametric-MNL: fi = β(si , xi ). Boosted trees: Partition the products into homogeneous groups in a way that respects the mean utility function.. Iteratively fits simple trees to explain errors not captured in the previous iteration. 8 / 16
  • 9. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Building Block: VC Trees Underlying VCM model: ξi = xi β(si ) + i , Piecewise constant approximation: ξi = M m=1 xi βmI(si ∈Cm) + i , M: number of partitions. {Cm}M m=1: a partition of the space of si . Piecewise constant approximation to the unknown high-dimensional function & data-driven partitioning method to obtain homogeneous regression relationships. Algorithm: Heuristics: greedy algorithm based on binary splits of the space of si (similar to CART). Splitting criterion: reduction in SSE. 9 / 16
  • 10. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Boosted VC-MNL Boosted VC-MNL: φ(f) = −2 K i=1 ni log(g(xi β(si ))) + 2N log K i=1 g(xi β(si )) + const. 1 Start with naive fit ˆf (0) = (x1β (0) , · · · , xK β (0) ) . 2 For b = 1, · · · , B, repeat: Compute the “pseudo observations”: ξi = − ∂φ ∂fi f =ˆf (b−1) . Fit ξi on si and xi using the “PartReg” algorithm to obtain partitions (C (b) 1 , · · · , C (b) M ). Let zi = (I (si ∈C (b) 1 ) , · · · , I (si ∈C (b) M ) , xi I (si ∈C (b) 1 ) , · · · , xi I (si ∈C (b) M ) ) , and use IRLS to estimate β (b) by minimizing J(β (b) ) = −2 K i=1 ni log(g(ˆf (b−1) i + zi β (b) )) + 2N log    K i=1 g(ˆf (b−1) i + zi β (b) )    . Update the fitted model by ˆf (b) = ˆf (b−1) + ν M m=1 β (b) 0m + β (b) 1mxi I (si ∈C (b) m ) . 3 Output the fitted model ˆf = ˆf (B). 10 / 16
  • 11. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Boosted VC-MNL Start with naive fit: e.g., simple linear MNL. Begin the iteration process: Compute pseudo observations/residuals. Fit an appropriate tree to predict pseudo residualts. Generate design matrix based on tree partitions, and fit linear MNL model. Addtive model of trees, not of predictors. Iteratively fit linear MNL models based on data-driven piecewise constant “bases”. 11 / 16
  • 12. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Mobile Computer Sales in Australia 6 months, 5 states; 30 choice sets (25 training, 5 test); use price residuals instead of price. Varying coefficient-MNL: fi = xi β(si ). Partially linear-MNL: fi = β0(si ) + xi β1. Nonparametric-MNL: fi = β(si , xi ). 0 200 400 600 800 1000 0.00.20.40.60.81.0 Varying coefficient−MNL, Boosted Iterations R2 Training Test 0 200 400 600 800 1000 0.00.20.40.60.81.0 Partially linear, Boosted Iterations R2 Training Test 0 200 400 600 800 1000 0.00.20.40.60.81.0 Nonparametric, Boosted Iterations R2 Training Test 12 / 16
  • 13. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Competitor Method – Elastic Net MNL Models: fi = xi β(si ). Linear-MNL: linear β(si ). Quadratic-MNL (first-order interaction). Quadratic-MNL: Initial features si . ⇒ Quadratic & first-order interaction among si , obtain design matrix zi . ⇒ Linear specification: β0(si ) = zi γ0 and β1(si ) = zi γ1. Elastic net (Zou & Hastie 2005) MNL: arg min γ0,γ1 −2 K i=1 ni log(g(zi γ0 + (zi xi )γ1)) + 2N log    K i=1 g(zi γ0 + (zi xi )γ1)    +λ    α i,j |γij | + (1 − α) 2 i,j γ 2 ij    α = 0: Ridge regression; α = 1: LASSO. g(·) : link function. Sparse and stable coefficient estimates, penalized IRLS. 13 / 16
  • 14. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Summary of Results Utility Optimal R2 Interactions Specification Estimation Training Test Time (min) Among attributes (α = 1) 399 .357 .17 X Linear (α = 1 2 ) .419 .379 .48 X (α = 1) penalized IRLS .582 .499 76.91 1st -order Quadratic (α = 1 2 ) .554 .53 52.78 1st -order Varying-coef. .734 .697 186.47 (B=1000) Partially linear boosted trees .493 .455 24.63 (B=1000) 2nd -order (M=4) Nonparametric .52 .502 23.43 (B=1000) M – size of each base tree; B– the number of boosting iterations Nonparametric MNL specifies a larger model space than VC-MNL, but piecewise constant trees fails to find the particular interactions. 14 / 16
  • 15. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Discussion Semiparametric MNL models, estimated by boosted tree methods. Learning from large-scale market data to a) make predictions and b) gain insights: econometrics & statistical learning. Statistical questions: Assessing errors in R2 and coefficient surface. Split selection in tree partitioning (variable importance). Model validation & diagnostics (standardized pseudo residuals). Choice of link functions. 15 / 16
  • 16. Motivation Aggregated Market Multinomial Logit Model Application to Australian Data Jianqiang (Jay) Wang Information Analytics Lab Hewlett-Packard Labs jianqiang.jay.wang@hp.com Thank you very much! 16 / 16