This document discusses algorithms for robotic dynamics and state estimation in uncertain environments. It is divided into two parts:
Part 1 covers deterministic dynamics and control of articulated rigid bodies using algorithms like the Newton-Euler recursive algorithm.
Part 2 discusses stochastic dynamics, including filtering, estimation and prediction algorithms like the Kalman filter that can estimate states from noisy sensor measurements by recursively applying Bayes' rule. It introduces concepts like the Gaussian prior and posterior distributions.
This document discusses various approaches for data fusion, which refers to statistically combining data from different sources. The main approaches covered are data assimilation, optimal interpolation, variational methods, and the Kalman filter. Data assimilation aims to combine model output with observations to estimate the true state. Optimal interpolation finds the best linear combination of a background field and observations to minimize error. Variational methods determine the state by minimizing a cost function, while the Kalman filter sequentially assimilates observations using forecast and analysis steps. The goal of all these approaches is to integrate multiple data sources to obtain a better estimate of the true state than using any one source alone.
In this talk we present a framework for splitting data assimilation problems based upon the model dynamics. This is motivated by assimilation in the unstable subspace (AUS) and center manifold and inertial manifold techniques in dynamical systems. Recent efforts based upon the development of particle filters projected into the unstable subspace will be highlighted.
Importance sampling has been widely used to improve the efficiency of deterministic computer simulations where the simulation output is uniquely determined, given a fixed input. To represent complex system behavior more realistically, however, stochastic computer models are gaining popularity. Unlike deterministic computer simulations, stochastic simulations produce different outputs even at the same input. This extra degree of stochasticity presents a challenge for reliability assessment in engineering system designs. Our study tackles this challenge by providing a computationally efficient method to estimate a system's reliability. Specifically, we derive the optimal importance sampling density and allocation procedure that minimize the variance of a reliability estimator. The application of our method to a computationally intensive, aeroelastic wind turbine simulator demonstrates the benefits of the proposed approaches.
In the first part, Tonda will give a brief introduction to quantum mechanics illustrated on simple discrete systems much like Feynman has in his famous lectures. We will focus on the concepts of spin and photon polarisation.
In the second part we will, one by one, tackle the basics of quantum computation: qubit, quantum computational operation, universal set of quantum instructions and how they relate to Turing machines.
And finally we will take a look at some of the most famous quantum algorithms and for some of them really look under the hood. Also, we will go through a brief overview of the experiments that have been run over the world in past decade or so.
This document defines and explains the properties of a contraction mapping. It states that a contraction mapping T on a complete metric space (X,d) satisfies d(T(x),T(y)) ≤ λd(x,y) for some 0 ≤ λ < 1. This ensures T has a unique fixed point x* where T(x*)=x*. The document proves convergence to the fixed point by showing the distance between successive terms xn+1=T(xn) goes to 0 as n goes to infinity. It also proves uniqueness of the fixed point by showing any other possible fixed point y* must equal x*.
Robust model predictive control for discrete-time fractional-order systemsPantelis Sopasakis
In this paper we propose a tube-based robust model predictive control scheme for fractional-order discrete-
time systems of the Grunwald-Letnikov type with state and input constraints. We first approximate the infinite-dimensional fractional-order system by a finite-dimensional linear system and we show that the actual dynamics can be approximated arbitrarily tight. We use the approximate dynamics to design a tube-based model predictive controller which endows to the controlled closed-loop system robust stability properties
Analysis of large scale spiking networks dynamics with spatio-temporal constr...Hassan Nasser
Recent experimental advances have made it possible to record up to several hundreds of neurons simultaneously in the cortex or in the retina. Analysing such data requires mathematical and numerical methods to describe the spatio-temporal correlations in population activity. This can be done thanks to Maximum Entropy method. Here, a crucial parameter is the product NxR where N is the number of neurons and R the memory depth of correlations (how far in the past does the spike activity affects the current state). Standard statistical mechanics methods are limited to spatial correlation structure with
R = 1 (e.g. Ising model) whereas methods based on transfer matrices, allowing the analysis of spatio-temporal correlations, are limited to NR = 20.
In the first part of the thesis we propose a modified version of the transfer matrix method, based on the parallel version of the Montecarlo algorithm, allowing us to go to NR = 100.
In the second part we present EnaS, a C++ library with a Graphical User Interface developed for neuroscientists. EnaS offers highly interactive tools that allow users to manage data, perform empirical statistics, modeling and visualizing results.
Finally, in a third part, we test our method on synthetic and real data sets. Real data set correspond to retina data provided by neuroscientists partners. Our non extensive analysis shows the advantages of considering spatio-temporal correlations for the analysis of retina spike trains, but it also outlines the limits of Maximum Entropy methods.
For more information about the software that I co-developed with my colleagues, please visit this page:
https://enas.inria.fr/
For more information about the publications, please visit this page:
https://scholar.google.fr/citations?user=L97ZODwAAAAJ
For the thesis, please visit this link:
https://www.theses.fr/178166669
How to Develop Your Own Simulators for Discrete-Event SystemsDonghun Kang
The document discusses how to develop simulators for discrete event systems using event graph and activity cycle diagram modeling approaches. It provides details on:
1. Functions for handling events, generating random variates, and event routines needed to simulate an event graph model.
2. The next event scheduling algorithm which proceeds by processing future events in time order and executing their associated event routines.
3. An example event graph model of a single server system and the pseudo-code for a simulator implementing the next event scheduling algorithm to execute this model.
This document discusses various approaches for data fusion, which refers to statistically combining data from different sources. The main approaches covered are data assimilation, optimal interpolation, variational methods, and the Kalman filter. Data assimilation aims to combine model output with observations to estimate the true state. Optimal interpolation finds the best linear combination of a background field and observations to minimize error. Variational methods determine the state by minimizing a cost function, while the Kalman filter sequentially assimilates observations using forecast and analysis steps. The goal of all these approaches is to integrate multiple data sources to obtain a better estimate of the true state than using any one source alone.
In this talk we present a framework for splitting data assimilation problems based upon the model dynamics. This is motivated by assimilation in the unstable subspace (AUS) and center manifold and inertial manifold techniques in dynamical systems. Recent efforts based upon the development of particle filters projected into the unstable subspace will be highlighted.
Importance sampling has been widely used to improve the efficiency of deterministic computer simulations where the simulation output is uniquely determined, given a fixed input. To represent complex system behavior more realistically, however, stochastic computer models are gaining popularity. Unlike deterministic computer simulations, stochastic simulations produce different outputs even at the same input. This extra degree of stochasticity presents a challenge for reliability assessment in engineering system designs. Our study tackles this challenge by providing a computationally efficient method to estimate a system's reliability. Specifically, we derive the optimal importance sampling density and allocation procedure that minimize the variance of a reliability estimator. The application of our method to a computationally intensive, aeroelastic wind turbine simulator demonstrates the benefits of the proposed approaches.
In the first part, Tonda will give a brief introduction to quantum mechanics illustrated on simple discrete systems much like Feynman has in his famous lectures. We will focus on the concepts of spin and photon polarisation.
In the second part we will, one by one, tackle the basics of quantum computation: qubit, quantum computational operation, universal set of quantum instructions and how they relate to Turing machines.
And finally we will take a look at some of the most famous quantum algorithms and for some of them really look under the hood. Also, we will go through a brief overview of the experiments that have been run over the world in past decade or so.
This document defines and explains the properties of a contraction mapping. It states that a contraction mapping T on a complete metric space (X,d) satisfies d(T(x),T(y)) ≤ λd(x,y) for some 0 ≤ λ < 1. This ensures T has a unique fixed point x* where T(x*)=x*. The document proves convergence to the fixed point by showing the distance between successive terms xn+1=T(xn) goes to 0 as n goes to infinity. It also proves uniqueness of the fixed point by showing any other possible fixed point y* must equal x*.
Robust model predictive control for discrete-time fractional-order systemsPantelis Sopasakis
In this paper we propose a tube-based robust model predictive control scheme for fractional-order discrete-
time systems of the Grunwald-Letnikov type with state and input constraints. We first approximate the infinite-dimensional fractional-order system by a finite-dimensional linear system and we show that the actual dynamics can be approximated arbitrarily tight. We use the approximate dynamics to design a tube-based model predictive controller which endows to the controlled closed-loop system robust stability properties
Analysis of large scale spiking networks dynamics with spatio-temporal constr...Hassan Nasser
Recent experimental advances have made it possible to record up to several hundreds of neurons simultaneously in the cortex or in the retina. Analysing such data requires mathematical and numerical methods to describe the spatio-temporal correlations in population activity. This can be done thanks to Maximum Entropy method. Here, a crucial parameter is the product NxR where N is the number of neurons and R the memory depth of correlations (how far in the past does the spike activity affects the current state). Standard statistical mechanics methods are limited to spatial correlation structure with
R = 1 (e.g. Ising model) whereas methods based on transfer matrices, allowing the analysis of spatio-temporal correlations, are limited to NR = 20.
In the first part of the thesis we propose a modified version of the transfer matrix method, based on the parallel version of the Montecarlo algorithm, allowing us to go to NR = 100.
In the second part we present EnaS, a C++ library with a Graphical User Interface developed for neuroscientists. EnaS offers highly interactive tools that allow users to manage data, perform empirical statistics, modeling and visualizing results.
Finally, in a third part, we test our method on synthetic and real data sets. Real data set correspond to retina data provided by neuroscientists partners. Our non extensive analysis shows the advantages of considering spatio-temporal correlations for the analysis of retina spike trains, but it also outlines the limits of Maximum Entropy methods.
For more information about the software that I co-developed with my colleagues, please visit this page:
https://enas.inria.fr/
For more information about the publications, please visit this page:
https://scholar.google.fr/citations?user=L97ZODwAAAAJ
For the thesis, please visit this link:
https://www.theses.fr/178166669
How to Develop Your Own Simulators for Discrete-Event SystemsDonghun Kang
The document discusses how to develop simulators for discrete event systems using event graph and activity cycle diagram modeling approaches. It provides details on:
1. Functions for handling events, generating random variates, and event routines needed to simulate an event graph model.
2. The next event scheduling algorithm which proceeds by processing future events in time order and executing their associated event routines.
3. An example event graph model of a single server system and the pseudo-code for a simulator implementing the next event scheduling algorithm to execute this model.
Supervisory control of discrete event systems for linear temporal logic speci...AmiSakakibara
This document provides an overview of a dissertation on supervisory control of discrete event systems for linear temporal logic specifications. It summarizes the chapters, including preliminaries on discrete event systems and supervisory control, an introduction to syntactically co-safe linear temporal logic, and an explanation of an on-line permissive supervisory control approach for scLTL specifications. This approach uses a ranking function to guide the supervisor toward accepting states and a permissiveness function to balance permissiveness over time.
Tracking involves predicting the motion of objects in a sequence of images to reduce computational costs compared to detecting objects in each frame independently. The Kalman filter provides a recursive method to estimate the state of a dynamic system and track objects over time based on sensor measurements and motion models. It minimizes the mean of the squared error between the estimated and actual state. Applications include structure from motion to jointly estimate camera motion and 3D scene structure from 2D image point correspondences over multiple frames.
This document presents a general framework for enhancing time series prediction performance. It discusses using multiple predictions from a base method like neural networks, ARIMA or Holt-Winters to improve accuracy. Short-term enhancement uses support vector regression on statistic and reliability features of the multiple predictions to enhance 1-step ahead predictions. Long-term enhancement trains additional models on the short-term predictions to enhance longer-horizon predictions. The framework is evaluated on traffic flow data with prediction horizons of 1 week and 13 weeks.
Stochastic augmentation by generalized minimum variance control with rst loop...UFPA
In this work we use the RST structure to shape the GMV optimization problem. The RST
controller is tuned by pole assignment based on a second order plant model and a second order desired closedloop model. The derived RST controller is then passed to the GMV generalized output weighting polynomials
in order to produce a stochastic equivalent controller. The result is the equivalence of the RST and the GMV
produced closed-loop dynamics whilst in ideal conditions (without noise or uncertainties), but a more economic
and efficient GMV closed-loop dynamics under an adverse stochastic scenario
Distributed solution of stochastic optimal control problem on GPUsPantelis Sopasakis
Stochastic optimal control problems arise in many
applications and are, in principle,
large-scale involving up to millions of decision variables. Their
applicability in control applications is often limited by the
availability of algorithms that can solve them efficiently and within
the sampling time of the controlled system.
In this paper we propose a dual accelerated proximal
gradient algorithm which is amenable to parallelization and
demonstrate that its GPU implementation affords high speed-up
values (with respect to a CPU implementation) and greatly outperforms
well-established commercial optimizers such as Gurobi.
The document discusses 2D transformations in computer graphics, including translation, scaling, rotation, and shearing. It provides details on how each transformation is represented by a matrix and how they can be combined by multiplying the matrices. Transformations allow repositioning, resizing, and changing the orientation of 2D objects. The key matrix for each transformation is:
- Translation: Adds translation offsets tx and ty
- Scaling: Scales by factors sx and sy
- Rotation: Rotates by an angle θ using cosine and sine functions
- Shearing: Shears the x-coordinates by a factor sh or y-coordinates by hs
Transformations can be combined by multiplying the matrices in sequence. The order
This document provides an overview of the Kalman filter, including its derivation and applications. It begins with an example of using a Kalman filter to estimate the position and velocity of a truck moving on rails. It then presents the general setup of the Kalman filter for estimating a stochastic dynamic system based on noisy observations. The derivation shows how to recursively calculate the optimal estimate and its error covariance. Several numerical examples are provided to illustrate Kalman filtering, including estimating a constant voltage, tracking a vehicle's position in 1D and 2D.
This document discusses recent advances in Markov chain Monte Carlo (MCMC) and sequential Monte Carlo (SMC) methods. It introduces Markov chain and sequential Monte Carlo techniques such as the Hastings-Metropolis algorithm, Gibbs sampling, data augmentation, and space alternating data augmentation. These techniques are applied to problems such as parameter estimation for finite mixtures of Gaussians.
1. The document discusses approaches for click-through rate (CTR) prediction, including offline and online methods as well as Bayesian parameter estimation.
2. It presents logistic regression models for CTR prediction with regularization to avoid overfitting, describing offline solutions using independent component optimization and online solutions using sample-by-sample updates.
3. Bayesian approaches are also covered, including maximizing the posterior probability to select hyperparameters like lambda and maximizing evidence to estimate model parameters.
A very wide spectrum of optimization problems can be efficiently solved with proximal gradient methods which hinge on the celebrated forward-backward splitting (FBS) schema. But such first-order methods are only effective when low or medium accuracy is required and are known to be rather slow or even impractical for badly conditioned problems. Moreover, the straightforward introduction of second-order (Hessian) information is beset with shortcomings as, typically, at every iteration we need to solve a non-separable optimisation problem. In this talk we will follow a different route to the solution of such optimisation problems. We will recast non-smooth optimisation problems as the minimisation of a real-valued, continuously differentiable function known as the forward-backward envelope. We will then employ a semismooth Newton method to solve the equivalent optimisation problem instead of the original one. We will then apply the proposed semismooth Newton method to L1-regularised least squares (LASSO) problems which is motivated by an an interesting application: recursive compressed sensing. Compressed sensing is a signal processing methodology for the reconstruction of sparsely sampled signals and it offers a new paradigm for sampling signals based on their innovation, that is, the minimum number of coefficients sufficient to accurately represent it in an appropriately selected basis. Compressed sensing leads to a lower sampling rate compared to theories using some fixed basis and has many applications in image processing, medical imaging and MRI, photography, holography, facial recognition, radio astronomy, radar technology and more. The traditional compressed sensing approach is naturally offline, in that it amounts to sparsely sampling and reconstructing a given dataset. Recently, an online algorithm for performing compressed sensing on streaming data was proposed; the scheme uses recursive sampling of the input stream and recursive decompression to accurately estimate stream entries from the acquired noisy measurements. We will see how we can tailor the forward-backward Newton method to solve recursive compressed sensing problems at one tenth of the time required by other algorithms such as ISTA, FISTA, ADMM and interior-point methods (L1LS).
Bayesian Inference : Kalman filter 에서 Optimization 까지 - 김홍배 박사님AI Robotics KR
[AI x Robotics : The First] 행사 - 김홍배 박사님 강연
Bayesian Inference : Kalman filter 에서 Optimization 까지
AI Robotics KR
(https://www.facebook.com/groups/airoboticskr/)
Debabrata Pal, Aksum University, College of Engineering and Technology Department of Electrical and Computer Engineering Ethiopia, NE Africa, Email:debuoisi@gmail.com,website:www.ijrd.in
We present a novel modeling
methodology to derive a nonlinear dynamical model which
adequately describes the effect of fuel sloshing on the attitude dynamics of a spacecraft. We model the impulsive thrusters using mixed logic and dynamics leading to a hybrid formulation.
We design a hybrid model predictive control scheme for the
attitude control of a launcher during its long coasting period,
aiming at minimising the actuation count of the thrusters.
The document describes the quicksort algorithm. Quicksort works by:
1) Partitioning the array around a pivot element into two sub-arrays of less than or equal and greater than elements.
2) Recursively sorting the two sub-arrays.
3) Combining the now sorted sub-arrays.
In the average case, quicksort runs in O(n log n) time due to balanced partitions at each recursion level. However, in the worst case of an already sorted input, it runs in O(n^2) time due to highly unbalanced partitions. A randomized version of quicksort chooses pivots randomly to avoid worst case behavior.
Phase Retrieval: Motivation and TechniquesVaibhav Dixit
This presentation describes two techniques namely Transport of Intensity Equation(TIE) technique and Phase Diversity technique for retrieving phase information from light.
Control of Discrete-Time Piecewise Affine Probabilistic Systems using Reachab...Leo Asselborn
This presentation proposes an algorithmic approach to
synthesize stabilizing control laws for discrete-time piecewise
affine probabilistic (PWAP) systems based on computations of
probabilistic reachable sets. The considered class of systems
contains probabilistic components (with Gaussian distribution)
modeling additive disturbances and state initialization. The
probabilistic reachable state sets contain all states that are
reachable with a given confidence level under the effect of
time-variant control laws. The control synthesis uses principles
of the ellipsoidal calculus, and it considers that the system
parametrization depends on the partition of the state space. The
proposed algorithm uses LMI-constrained semi-definite programming
(SDP) problems to compute stabilizing controllers,
while polytopic input constraints and transitions between regions
of the state space are considered. The formulation of
the SDP is adopted from a previous work in [1] for switched
systems, in which the switching of the continuous dynamics
is triggered by a discrete input variable. Here, as opposed
to [1], the switching occurs autonomously and an algorithmic
procedure is suggested to synthesis a stabilizing controller. An
example for illustration is included.
Control of Uncertain Nonlinear Systems Using Ellipsoidal Reachability CalculusLeo Asselborn
This paper proposes an approach to algorithmically synthesize control strategies for discrete-time nonlinear uncertain systems based on reachable set computations using the ellipsoidal calculus. For given ellipsoidal initial sets and bounded ellipsoidal disturbances, the proposed algorithm iterates over conservatively approximating and LMI-constrained optimization problems to compute stabilizing controllers. The method uses
first-order Taylor approximation of the nonlinear dynamics and a conservative approximation of the Lagrange remainder.
Differential Geometry for Machine LearningSEMINARGROOT
References:
Differential Geometry of Curves and Surfaces, Manfredo P. Do Carmo (2016)
Differential Geometry by Claudio Arezzo
Youtube: https://youtu.be/tKnBj7B2PSg
What is a Manifold?
Youtube: https://youtu.be/CEXSSz0gZI4
Shape analysis (MIT spring 2019) by Justin Solomon
Youtube: https://youtu.be/GEljqHZb30c
Tensor Calculus
Youtube: https://youtu.be/kGXr1SF3WmA
Manifolds: A Gentle Introduction,
Hyperbolic Geometry and Poincaré Embeddings by Brian Keng
Link: http://bjlkeng.github.io/posts/manifolds/,
http://bjlkeng.github.io/posts/hyperbolic-geometry-and-poincare-embeddings/
Statistical Learning models for Manifold-Valued measurements with application to computer vision and neuroimaging by Hyunwoo J.Kim
MVPA with SpaceNet: sparse structured priorsElvis DOHMATOB
The GraphNet (aka S-Lasso), as well as other “sparsity + structure” priors like TV (Total-Variation), TV-L1, etc., are not easily applicable to brain data because of technical problems
relating to the selection of the regularization parameters. Also, in
their own right, such models lead to challenging high-dimensional optimization problems. In this manuscript, we present some heuristics for speeding up the overall optimization process: (a) Early-stopping, whereby one halts the optimization process when the test score (performance on leftout data) for the internal cross-validation for model-selection stops improving, and (b) univariate feature-screening, whereby irrelevant (non-predictive) voxels are detected and eliminated before the optimization problem is entered, thus reducing the size of the problem. Empirical results with GraphNet on real MRI (Magnetic Resonance Imaging) datasets indicate that these heuristics are a win-win strategy, as they add speed without sacrificing the quality of the predictions. We expect the proposed heuristics to work on other models like TV-L1, etc.
The issues about maneuvering target track prediction were discussed in this paper. Firstly, using Kalman filter which based on current statistical model describes the state of maneuvering target motion, thereby analyzing time range of the target maneuvering occurred. Then, predict the target trajectory in real time by the improved gray prediction model. Finally, residual test and posterior variance test model accuracy, model accuracy is accurate.
Supervisory control of discrete event systems for linear temporal logic speci...AmiSakakibara
This document provides an overview of a dissertation on supervisory control of discrete event systems for linear temporal logic specifications. It summarizes the chapters, including preliminaries on discrete event systems and supervisory control, an introduction to syntactically co-safe linear temporal logic, and an explanation of an on-line permissive supervisory control approach for scLTL specifications. This approach uses a ranking function to guide the supervisor toward accepting states and a permissiveness function to balance permissiveness over time.
Tracking involves predicting the motion of objects in a sequence of images to reduce computational costs compared to detecting objects in each frame independently. The Kalman filter provides a recursive method to estimate the state of a dynamic system and track objects over time based on sensor measurements and motion models. It minimizes the mean of the squared error between the estimated and actual state. Applications include structure from motion to jointly estimate camera motion and 3D scene structure from 2D image point correspondences over multiple frames.
This document presents a general framework for enhancing time series prediction performance. It discusses using multiple predictions from a base method like neural networks, ARIMA or Holt-Winters to improve accuracy. Short-term enhancement uses support vector regression on statistic and reliability features of the multiple predictions to enhance 1-step ahead predictions. Long-term enhancement trains additional models on the short-term predictions to enhance longer-horizon predictions. The framework is evaluated on traffic flow data with prediction horizons of 1 week and 13 weeks.
Stochastic augmentation by generalized minimum variance control with rst loop...UFPA
In this work we use the RST structure to shape the GMV optimization problem. The RST
controller is tuned by pole assignment based on a second order plant model and a second order desired closedloop model. The derived RST controller is then passed to the GMV generalized output weighting polynomials
in order to produce a stochastic equivalent controller. The result is the equivalence of the RST and the GMV
produced closed-loop dynamics whilst in ideal conditions (without noise or uncertainties), but a more economic
and efficient GMV closed-loop dynamics under an adverse stochastic scenario
Distributed solution of stochastic optimal control problem on GPUsPantelis Sopasakis
Stochastic optimal control problems arise in many
applications and are, in principle,
large-scale involving up to millions of decision variables. Their
applicability in control applications is often limited by the
availability of algorithms that can solve them efficiently and within
the sampling time of the controlled system.
In this paper we propose a dual accelerated proximal
gradient algorithm which is amenable to parallelization and
demonstrate that its GPU implementation affords high speed-up
values (with respect to a CPU implementation) and greatly outperforms
well-established commercial optimizers such as Gurobi.
The document discusses 2D transformations in computer graphics, including translation, scaling, rotation, and shearing. It provides details on how each transformation is represented by a matrix and how they can be combined by multiplying the matrices. Transformations allow repositioning, resizing, and changing the orientation of 2D objects. The key matrix for each transformation is:
- Translation: Adds translation offsets tx and ty
- Scaling: Scales by factors sx and sy
- Rotation: Rotates by an angle θ using cosine and sine functions
- Shearing: Shears the x-coordinates by a factor sh or y-coordinates by hs
Transformations can be combined by multiplying the matrices in sequence. The order
This document provides an overview of the Kalman filter, including its derivation and applications. It begins with an example of using a Kalman filter to estimate the position and velocity of a truck moving on rails. It then presents the general setup of the Kalman filter for estimating a stochastic dynamic system based on noisy observations. The derivation shows how to recursively calculate the optimal estimate and its error covariance. Several numerical examples are provided to illustrate Kalman filtering, including estimating a constant voltage, tracking a vehicle's position in 1D and 2D.
This document discusses recent advances in Markov chain Monte Carlo (MCMC) and sequential Monte Carlo (SMC) methods. It introduces Markov chain and sequential Monte Carlo techniques such as the Hastings-Metropolis algorithm, Gibbs sampling, data augmentation, and space alternating data augmentation. These techniques are applied to problems such as parameter estimation for finite mixtures of Gaussians.
1. The document discusses approaches for click-through rate (CTR) prediction, including offline and online methods as well as Bayesian parameter estimation.
2. It presents logistic regression models for CTR prediction with regularization to avoid overfitting, describing offline solutions using independent component optimization and online solutions using sample-by-sample updates.
3. Bayesian approaches are also covered, including maximizing the posterior probability to select hyperparameters like lambda and maximizing evidence to estimate model parameters.
A very wide spectrum of optimization problems can be efficiently solved with proximal gradient methods which hinge on the celebrated forward-backward splitting (FBS) schema. But such first-order methods are only effective when low or medium accuracy is required and are known to be rather slow or even impractical for badly conditioned problems. Moreover, the straightforward introduction of second-order (Hessian) information is beset with shortcomings as, typically, at every iteration we need to solve a non-separable optimisation problem. In this talk we will follow a different route to the solution of such optimisation problems. We will recast non-smooth optimisation problems as the minimisation of a real-valued, continuously differentiable function known as the forward-backward envelope. We will then employ a semismooth Newton method to solve the equivalent optimisation problem instead of the original one. We will then apply the proposed semismooth Newton method to L1-regularised least squares (LASSO) problems which is motivated by an an interesting application: recursive compressed sensing. Compressed sensing is a signal processing methodology for the reconstruction of sparsely sampled signals and it offers a new paradigm for sampling signals based on their innovation, that is, the minimum number of coefficients sufficient to accurately represent it in an appropriately selected basis. Compressed sensing leads to a lower sampling rate compared to theories using some fixed basis and has many applications in image processing, medical imaging and MRI, photography, holography, facial recognition, radio astronomy, radar technology and more. The traditional compressed sensing approach is naturally offline, in that it amounts to sparsely sampling and reconstructing a given dataset. Recently, an online algorithm for performing compressed sensing on streaming data was proposed; the scheme uses recursive sampling of the input stream and recursive decompression to accurately estimate stream entries from the acquired noisy measurements. We will see how we can tailor the forward-backward Newton method to solve recursive compressed sensing problems at one tenth of the time required by other algorithms such as ISTA, FISTA, ADMM and interior-point methods (L1LS).
Bayesian Inference : Kalman filter 에서 Optimization 까지 - 김홍배 박사님AI Robotics KR
[AI x Robotics : The First] 행사 - 김홍배 박사님 강연
Bayesian Inference : Kalman filter 에서 Optimization 까지
AI Robotics KR
(https://www.facebook.com/groups/airoboticskr/)
Debabrata Pal, Aksum University, College of Engineering and Technology Department of Electrical and Computer Engineering Ethiopia, NE Africa, Email:debuoisi@gmail.com,website:www.ijrd.in
We present a novel modeling
methodology to derive a nonlinear dynamical model which
adequately describes the effect of fuel sloshing on the attitude dynamics of a spacecraft. We model the impulsive thrusters using mixed logic and dynamics leading to a hybrid formulation.
We design a hybrid model predictive control scheme for the
attitude control of a launcher during its long coasting period,
aiming at minimising the actuation count of the thrusters.
The document describes the quicksort algorithm. Quicksort works by:
1) Partitioning the array around a pivot element into two sub-arrays of less than or equal and greater than elements.
2) Recursively sorting the two sub-arrays.
3) Combining the now sorted sub-arrays.
In the average case, quicksort runs in O(n log n) time due to balanced partitions at each recursion level. However, in the worst case of an already sorted input, it runs in O(n^2) time due to highly unbalanced partitions. A randomized version of quicksort chooses pivots randomly to avoid worst case behavior.
Phase Retrieval: Motivation and TechniquesVaibhav Dixit
This presentation describes two techniques namely Transport of Intensity Equation(TIE) technique and Phase Diversity technique for retrieving phase information from light.
Control of Discrete-Time Piecewise Affine Probabilistic Systems using Reachab...Leo Asselborn
This presentation proposes an algorithmic approach to
synthesize stabilizing control laws for discrete-time piecewise
affine probabilistic (PWAP) systems based on computations of
probabilistic reachable sets. The considered class of systems
contains probabilistic components (with Gaussian distribution)
modeling additive disturbances and state initialization. The
probabilistic reachable state sets contain all states that are
reachable with a given confidence level under the effect of
time-variant control laws. The control synthesis uses principles
of the ellipsoidal calculus, and it considers that the system
parametrization depends on the partition of the state space. The
proposed algorithm uses LMI-constrained semi-definite programming
(SDP) problems to compute stabilizing controllers,
while polytopic input constraints and transitions between regions
of the state space are considered. The formulation of
the SDP is adopted from a previous work in [1] for switched
systems, in which the switching of the continuous dynamics
is triggered by a discrete input variable. Here, as opposed
to [1], the switching occurs autonomously and an algorithmic
procedure is suggested to synthesis a stabilizing controller. An
example for illustration is included.
Control of Uncertain Nonlinear Systems Using Ellipsoidal Reachability CalculusLeo Asselborn
This paper proposes an approach to algorithmically synthesize control strategies for discrete-time nonlinear uncertain systems based on reachable set computations using the ellipsoidal calculus. For given ellipsoidal initial sets and bounded ellipsoidal disturbances, the proposed algorithm iterates over conservatively approximating and LMI-constrained optimization problems to compute stabilizing controllers. The method uses
first-order Taylor approximation of the nonlinear dynamics and a conservative approximation of the Lagrange remainder.
Differential Geometry for Machine LearningSEMINARGROOT
References:
Differential Geometry of Curves and Surfaces, Manfredo P. Do Carmo (2016)
Differential Geometry by Claudio Arezzo
Youtube: https://youtu.be/tKnBj7B2PSg
What is a Manifold?
Youtube: https://youtu.be/CEXSSz0gZI4
Shape analysis (MIT spring 2019) by Justin Solomon
Youtube: https://youtu.be/GEljqHZb30c
Tensor Calculus
Youtube: https://youtu.be/kGXr1SF3WmA
Manifolds: A Gentle Introduction,
Hyperbolic Geometry and Poincaré Embeddings by Brian Keng
Link: http://bjlkeng.github.io/posts/manifolds/,
http://bjlkeng.github.io/posts/hyperbolic-geometry-and-poincare-embeddings/
Statistical Learning models for Manifold-Valued measurements with application to computer vision and neuroimaging by Hyunwoo J.Kim
MVPA with SpaceNet: sparse structured priorsElvis DOHMATOB
The GraphNet (aka S-Lasso), as well as other “sparsity + structure” priors like TV (Total-Variation), TV-L1, etc., are not easily applicable to brain data because of technical problems
relating to the selection of the regularization parameters. Also, in
their own right, such models lead to challenging high-dimensional optimization problems. In this manuscript, we present some heuristics for speeding up the overall optimization process: (a) Early-stopping, whereby one halts the optimization process when the test score (performance on leftout data) for the internal cross-validation for model-selection stops improving, and (b) univariate feature-screening, whereby irrelevant (non-predictive) voxels are detected and eliminated before the optimization problem is entered, thus reducing the size of the problem. Empirical results with GraphNet on real MRI (Magnetic Resonance Imaging) datasets indicate that these heuristics are a win-win strategy, as they add speed without sacrificing the quality of the predictions. We expect the proposed heuristics to work on other models like TV-L1, etc.
The issues about maneuvering target track prediction were discussed in this paper. Firstly, using Kalman filter which based on current statistical model describes the state of maneuvering target motion, thereby analyzing time range of the target maneuvering occurred. Then, predict the target trajectory in real time by the improved gray prediction model. Finally, residual test and posterior variance test model accuracy, model accuracy is accurate.
This document provides a summary of key concepts in computed tomography (CT) imaging. It discusses back projection reconstruction techniques which can produce blurred images. It also describes filtered back projection which uses digital subtraction filtering to reduce blurring. Fourier reconstruction techniques are described where projection data is transformed to the frequency domain using Fourier transforms before being reconstructed into a spatial domain image. Different window widths and window levels are discussed for optimizing soft tissue versus bone imaging.
Introduction to Neural Networks and Deep Learning from ScratchAhmed BESBES
If you're willing to understand how neural networks work behind the scene and debug the back-propagation algorithm step by step by yourself, this presentation should be a good starting point.
We'll cover elements on:
- the popularity of neural networks and their applications
- the artificial neuron and the analogy with the biological one
- the perceptron
- the architecture of multi-layer perceptrons
- loss functions
- activation functions
- the gradient descent algorithm
At the end, there will be an implementation FROM SCRATCH of a fully functioning neural net.
code: https://github.com/ahmedbesbes/Neural-Network-from-scratch
1. The document provides instructions to solve problems related to digital waveguide oscillators, digital lattice filters, and other discrete-time linear systems. Students are asked to write state space equations, find eigenvalues, compute responses, and represent systems using different forms such as state space and block diagrams. MATLAB code is provided to help with computations.
2. Students must analyze cascaded and parallel systems, check controllability and observability, and represent pulse transfer functions using state space, direct form, cascade form, and other block diagram representations. They are also asked to transform state space representations between different coordinate systems.
The document describes a control framework called the "stack of tasks" which provides hierarchical task-based control for real-time redundant manipulators. It allows implementation of a data flow graph controlled by Python scripting. Tasks are defined as functions of the robot configuration, time, and other parameters that should converge to zero. The framework computes joint velocities to minimize higher priority tasks while satisfying lower priority tasks when possible. It has been tested on robots including HRP-2, Nao, and Romeo.
Pole placement by er. sanyam s. saini (me reg)Sanyam Singh
The document discusses state feedback control with integral control. It explains that state feedback alone does not provide integral control, so an extra integral state must be augmented to the plant model. This augmented model is then used to design state feedback control with an additional integral term that drives the output to the reference value and eliminates steady state error. An example is provided to illustrate designing such an integral control system for a given plant model with a constant reference input.
Wang-Landau Monte Carlo simulation is a method for calculating the density of states function which can then be used to calculate thermodynamic properties like the mean value of variables. It improves on traditional Monte Carlo methods which struggle at low temperatures due to complicated energy landscapes with many local minima separated by large barriers. The Wang-Landau algorithm calculates the density of states function directly rather than relying on sampling configurations, allowing it to overcome barriers and fully explore the configuration space even at low temperatures.
Digital Signal Processing[ECEG-3171]-Ch1_L03Rediet Moges
This Digital Signal Processing Lecture material is the property of the author (Rediet M.) . It is not for publication,nor is it to be sold or reproduced.
#Africa#Ethiopia
The document contains 24 sample questions for an AM paper. The questions cover topics related to binary, logic, computing hardware, algorithms and data structures. They include multiple choice questions testing knowledge of binary representations, logic expressions, computer architecture concepts like cache memory and pipelining, and algorithms involving queues, trees and arrays.
This document discusses object detection using Adaboost and various techniques. It begins with an overview of the Adaboost algorithm and provides a toy example to illustrate how it works. Next, it describes how Viola and Jones used Adaboost with Haar-like features and an integral image representation for rapid face detection in images. It achieved high detection rates with very low false positives. The document also discusses how Schneiderman and Kanade used a parts-based representation with localized wavelet coefficients as features for object detection and used statistical independence of parts to obtain likelihoods for classification.
This document describes the design of a servo system using state feedback and integral control. It defines the plant state and output equations, and shows the block diagram of the servo system. The state equation of the augmented system is derived, combining the plant states and integrator states. The gains K1 and K2 are selected using pole placement so that the closed-loop poles of the combined system are located at the desired locations. An example is provided to illustrate the design process.
1) The document describes a fractional order nonlinear quarter car suspension model. It establishes integer and fractional order differential equations to model the system.
2) Key parameters of the suspension system are defined including mass, stiffness coefficients, and hysteretic nonlinear damping forces. State space and discrete forms of the fractional order model are presented.
3) Numerical methods for solving the fractional order differential equations are discussed, including the Adams-Bashforth-Moulton algorithm used to analyze the quarter car model. Stability of equilibrium points is analyzed.
The document discusses the divide and conquer algorithm design strategy. It begins by explaining the general concept of divide and conquer, which involves splitting a problem into subproblems, solving the subproblems, and combining the solutions. It then provides pseudocode for a generic divide and conquer algorithm. Finally, it gives examples of divide and conquer algorithms like quicksort, binary search, and matrix multiplication.
AU QP Answer key NOv/Dec 2015 Computer Graphics 5 sem CSEThiyagarajan G
This document contains a summary of a computer graphics exam with 10 multiple choice questions in Part A and 4 long answer questions in Part B. Some of the key topics covered include: image resolution, scaling matrices, color conversion between RGB and CMY color modes, Bezier curves, projection planes, dithering, animation principles, turtle attributes in graphics, Bresenham's circle algorithm, Liang-Barsky line clipping algorithm, viewing transformations, cubic Bezier curves, and backface detection. Part B also includes questions on orthographic vs axonometric vs oblique projections, ambient lighting models, raster vs keyframe animation, ray tracing, and morphing.
This document contains a homework assignment for a 4th year biomedical engineering systems engineering course. It involves solving several problems related to modeling and analyzing dynamic systems using MATLAB. Students are asked to write individual reports for each problem and not copy solutions from classmates. The problems cover topics like determining states of linear systems, showing equivalence of different representations, analyzing controllability and observability, designing state feedback gains, and developing an integral control system. MATLAB functions for obtaining state feedback and integral gains are also to be implemented.
The document discusses various load forecasting methods used in power systems, including:
1) Exponential smoothing techniques like linear, exponential, and polynomial regression to model load growth over time.
2) Land use simulation to map existing and planned development to forecast load growth.
3) Box-Jenkins methodology using autoregressive and moving average processes to model load patterns for short-term forecasting.
Conversion of transfer function to canonical state variable modelsJyoti Singh
Realization of transfer function into state variable models is needed even if the control system design based on frequency-domain design method.
In these cases the need arises for the purpose of transient response simulation.
But there is not much software for the numerical inversion of Laplace transform.
So one ways is to convert transfer function of the system to state variable description and numerically integrating the resulting differential equations rather than attempting to compute the inverse Laplace transform by numerical method.
The document provides a course calendar for a class on Bayesian estimation methods. It lists the dates and topics to be covered over 15 class periods from September to January. The topics progress from basic concepts like Bayes estimation and the Kalman filter, to more modern methods like particle filters, hidden Markov models, Bayesian decision theory, and applications of principal component analysis and independent component analysis. One class is noted as having no class.
3. Tonight's Recurring Themes
TranslationRotation
[Algorithms for computing NewtonianEulerian
kinematics]
ForwardBackward
[Algorithms for computing kinematicsdynamics
and state predictionestimation]
F=maAx=b
[Algorithms for predicting dynamics and estimating
states from measurements]
5. Some Definitions
Spatial Motion of Rigid Bodies
Denavit – Hartenberg Representation
Sheth - Uicker Representation
Computations and Algorithms
Kinematics of Some Wheeled Robots
Time Derivatives in ICC
The Wheel Jacobian
Computations and Algorithms
Velocity – Torque Duality
Newton—Euler Recursive Algorithm
Robot Control
Content of this Talk : Part 1 – Articulated Rigid Bodies
6. “In theory, there is no difference between
theory and practice. In practice there is.”
8. In the Beginning - Geometry of Ax=b
Minimizing Error - (Vanilla LS)
Weighted - (Vanilla w/ Cream LS)
Recursive - (Vanilla w /Cream & Nuts on top LS)
… and Beyond (The Kalman Filter)
Rules of Probability (Bayes’ POV)
Graphical Models - Deconstructing Bayes
Noisy Measurements and Estimation
Repeated Noisy Measurements - Recursive Bayes
Noisy Measurements and Estimation
Hidden Markov Model (HMM)
Forward—Backward Algorithm
Content of this Talk – Part 2 : Filtering, Estimation and Prediction
9. It's tough to make predictions, especially
about the future.
Yogi Berra
10. Gauss-Markov-Kalman [and sometimes ]
Carl Friederich
Gauss
1777-1855
Andrey Markov
1856-1922
Rudolf Kalman
1930
John Flaig
Flaig
Thomas Bayes
1701-1761
11. Motivation
Deterministic models represent adequate description
of dynamics and control – why complicate things with
stochasticity?
Incomplete Deterministic Models: Models are based
on assumptions and hence are approximations –
ignoring higher modes, does not make them go away.
Extraneous Disturbance: Systems are driven not just
by deterministic control inputs but also by uncontrollable
environmental factors – wind gusts, treacherous terrain.
Incomplete /Noisy Measurement: Not everything is
amenable to measurement (easier to measure position
than velocity). Measurement errors are unavoidable.
Estimate the state xk from
noisy measurements zk
12. How to …
Model Development: Develop models that account for
uncertainties that are practical to implement.
Optimal Estimation: How to estimate model behavior based on
incomplete and noisy sensor data – fuse data from multiple sources,
recursively in real-time.
Optimal Control: Given uncertain system description, incomplete,
noisy and corrupted data, how to optimally control a system for a
desirable performance.
Estimate the state xk from
noisy measurements zk
Performance Evaluation: How to evaluate performance capabilities
of estimation and control both before and after they are built.
13. Problem Formulation
System Dynamics :
Robot EOM : )(),()( qGqqVqqM
))(,),(),(),(()( tttutxtxftx
Observation : ))(,),(),(),(()( tttutxtxhtz
Linear Model :
)()()(
)()()()(
tvtHxtz
twtButAxtx
Discrete Time :
kkk
kkkk
vHxz
wBuAxx
111
k
k
z
x State @ tk
Observation @ tk
Estimate the state xk from
noisy measurements zk
14. Intro to Least Squares – Geometry of Ax=b
2
1
2
22
12
1
21
11
2
1
2
1
2221
1211
b
b
x
a
a
x
a
a
b
b
x
x
aa
aa
bAx
Linear Combination
of Columns of A
What if b is NOT in S(A)?
Solution exists if b lies in S(A)
[space spanned by columns of A]
Columns Space of A
bAx
Columns Space of A
PbxA ˆ
b
xAb ˆ
Project b on S(A) and call it
the best estimate of x.
xAbe ˆMinimize
15. bAAAxbAAxA TTTT 1
)(
Recall for a non-square A
Minimizing the Error – Vanilla LS
bAAAx
xAAbAe
xd
d
xAbxAbe
xAbe
TT
TT
T
1
2
2
)(ˆ
0ˆ22
ˆ
)ˆ).(ˆ(
ˆ
Projection P
(on S(A))
Minimize
Consider
bAx
The best solution is the one that minimizes the norm square of the error
(Assume b to be measurements and x the state)
16. WbWAWAWAx
WbWAxWAWA
ewewewWe
TTTT
TT
1
2
33
2
22
2
11
2
)(ˆ
)(ˆ)(
...
bAAAxbAAxA TTTT 1
)(
Recall when equally reliable W=I
dxxxpeE )(][
Mean Error
dxxpxeE )(][ 222
Variance
dxeepeeeeEcv jijiji ),(][
C0-Variance
1
V 1
V
][
)(ˆ
1
111
T
T
P
T
eeEP
bVAAVAx
Weighted Residual – Vanilla w/ Cream LS
If the measurements are not equally reliable
WbWAx (W is diagonal
of weights wi)
17. Recursive LS – Vanilla w/ Cream & Nuts on-the-top LS
Now imagine the data coming in a stream
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
0b
00 bxA
11 bxA
1b
If more data arrives, can the best estimate for
the combined data be computed from x0 and
b1 without restarting the calculation from b0?
Lets compute the
average of n numbers n
xxx
nA n...
)( 21
One additional
data arrives 1
...
)1( 121
n
xxxx
nA nn
121 ...)1()1( nn xxxxnAn
1211 ...)( nnn xxxxxnnA
)(
1
1
)()1( 1 nAx
n
nAnA n
Re-arrange
and simplify
)()()1( 1 nAxKnAnA n
Running Average
)()1(, nAnAn
A(n+1)
Digression – Running Average
18. Recursive LS – Vanilla w/ Cream & Nuts on-the-top LS …
2b
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
0b
00 bxA
11 bxA
1b
22 bxA
Appended
Data
1
0
1
0
1
1
0
1
0
b
b
A
A
x
A
A
A
A
TT
)( 11001
1
0
1
0
11 bAbAP
b
b
A
A
Px TT
T
Re-arrange
and simplify )( 0111101 xAbAPxx T
Original Data 00000 )( bAxAA
TT
)( 011101 xAbKxx
Recursive LS
19. Recursive LS – What’s Goin’ On?
Projecting b on S(A) is the
best estimate of x.
Columns Space of A
PbxA ˆ
b
xAb ˆ
exAb ˆError
Expanding th Columns
Space of A
bAppending data expands S(A)
and makes b closer to S(A) which
reduces the error. In the limit, b
collapses on S(A) we have the
exact Ax=b.
0ˆ xAbError
20. Recursive LS – and Beyond (Take One)
iii
iii
xAb
xFx
1 System Dynamics
Measurement
EstimateCurrent
State
Predict Future State
1b
1P
111 xAb
2
1
0
2
1
0
2
1
1
0
0
0
0
00
0
00
0
00
b
b
b
x
x
x
A
IF
A
IF
A
bAx )(1 iiiiii xAbKxx
Kalman
1x0F
100 xxF
0x
0b
0P
000 xAb
2b 3b
2x 3x 4x
2P 3P
1F 2F 3F
222 xAb
211 xxF 322 xxF 433 xxF
21. Recursive LS – and Beyond (Take Two)
)ˆ(ˆˆ
kkkkk xHzKxx
Kalman
kkk
kkk
vHxz
wAxx
11 System Dynamics with noise
Measurement with error
System &
Measurement Noise
kx
kz
kxˆ
1kx kx
ˆ
RRNvp
QQNwp
:),0(~)(
:),0(~)( system noise covariance
meas. noise covariance
0
ˆ
:][:ˆ
][:ˆ
k
kT
kkkkkk
T
kkkkkk
xd
dP
eeEPxxe
eeEPxxe
22. Rules of Probability – The Bayesian POV
)()|()()|(),(
),()(;),()(
ypyxpxpxypyxp
yxpypyxpxp
xy
Sum Rule
Product Rule
Joint Probability
Conditional Probability
)(
)()|(
)|(
xp
ypyxp
xyp
Bayes’ Rule
(Posterior)
(Likelihood) (Prior)
(Marginal)
(Prior) – belief before making and observation or collecting data.
(Posterior) – belief after making and observation or collecting
data. It forms the Prior for the next iteration.
(Likelihood) – it is a function of y, not a probability distribution
of y.
(Marginal) – Data from the observation.
p(x,y)
p(y|x)
p(x|y)
x
y
23. Rules of Probability – The Bayesian POV
(Posterior) )(
)()|(
)|(
xp
pxp
xp
Bayes’ Rule
(Likelihood) (Prior)
(Marginal)
2
,)|( Nxp Likelihood
Unknown Known
2
00 ,)( Np Prior (Known)
),().,()()|()|(
2
00
2
NNpxpxp
Posterior
Gaussian Prior
Gaussian Prior -> Gaussian Posterior
[Conjugate Pair]
24. Conditional Probability - De-Constructing Bayes’
),|().|(
),|().,,|(
).().().(
54746
2153214
321
xxxpxxp
xxxpxxxxp
xpxpxp
)().|().,|(),,( cpcbpcbapcbap
c
b
a
)|().|().(),,( bcpabpapcbap
cba
)().().(),,( cpbpapcbap
cba
26. Making Noisy Measurements on a Stationary Robot
Image Recognition/ Data
Display / Storage
Robot
Target
Camera Problem Statement
We want to estimate the mean position
() of the robot from the noisy image
measurements (x).
)(p
Measurement Model
Assume () to be normally
distributed random variable
with probability p().
x
(x) is the noisy measurement.
)|( xpNoise on the sensor
centered around (x) is the
noisy measurement.
Choose () .
27. 100um
Ext
PRE
Theta
-4
-3
-2
-1
0
1
2
3
Prin2
1
23
4
5
6
xy
-10 -8 -6 -4 -2 0 2 4
Prin 1
Extension
Pre-aligner
Theta
Theta
25
35
45
55
65
75
Meanofr
33 66 99 132 165 198 231 264 297 330 363 396 429 462
Sample 40 50 60
Wafer Placement
“Potential” of the System
Patterns in data indicate presence of assignable cause(s).
(Gaussian = Purely Random)
Pattern vs Noise – One More Digression …
30. Making Noisy Measurements on a Moving Robot (The Double Whammy)
WMR whose distance from an obstacle is
measured by an ultrasonic sensor. Estimate
the position of WMR at any time t.
Measurement: @ t=0, initial position has a
Gaussian distribution based on sensor
accuracy. Initial position is estimated.
Prediction: @ t=1, position is predicted from
the estimate @ t=0.
Measurement: @ t=1, position is estimated
from the measurement @ t=1.
Correction: @ t=1, position is corrected
from the measurement @ t=1. Also
prediction is made for t=2.
31. Making Noisy Measurements on a Moving Robot
System Noise – Low
Measurement Noise - Low
System Noise – Low
Measurement Noise - High
System Noise – High
Measurement Noise - Low
System Noise – High
Measurement Noise - High
32. Making Noisy Measurements on a Moving Robot (Take Three)
Kalman
iiii
iiii
exAb
xFx
1
x
)(p
)|( xp
1t 2t 3t 4t
)|( xp
Measurement
Noise
Prediction
Error
33. The HMM – Single Page Review
Hidden Markov Model
Hidden
States
Observations
Transition
T(i,j)
Xxnz kk };...1{
Emission
i
1t
1z 2z 3z kz 1kz
1x 2x 3x kx 1kx
2t 3t kt 1 kt
)( 1zp
k
kkkk zxpzzpzxpzpxzp )|()|()|()(),( 1111
Xxizxxpx
njiizjzpjiT
kkii
kk
),|()(
}..1{),();|(),( 1
35. The HMM – Example – 2 State Model
}10...1{};2,1{ Xxz kk (Uniformly Distributed)
7
2
9
1
6
2
7
1
6
1
36. Forward-Backward – Two Ways
3j mj2j1j
0jRobot
Dynamic
&
Control
(Featherstone) kkk maf .
Velocities, Accelerations
Forces, Torques
Prediction, Estimation
Smoothing
State
Prediction
&
Estimation
(Kalman)
(Space)
z
x
)(xp
)|( xzp
)|( zxp
(Time)
37. The Definition - Reliability of a Robot
It is the probability (R) that the robot will successfully complete
the assigned task (T) under the specified conditions (C).
Specified Conditions (C):
Martian Terrain / Contact with Human Body / Assembly Line
Assigned Task (T):
Move from A to B / Perform Surgery / Spot Weld.
Probability (R).
On Mars, move from A to B 50 times without failure.
On a human perform surgery with failure.
On an assembly line do 1 million spot welds before failure.
The basic problem is to quantify R during design.
39. References
1. Craig, J. J., Introduction to Robotics,: Mechanics and Control,
Prince-Hall, 2003.
2. Muir, F. P. and Neuman, C. P. , (1987), Kinematics Modeling of
Wheeled Mobile Robots, J. Robotic Systems, 4(2).
3. Bishop, C. , Pattern Recognition and Machine Learning, Spring ,
2006.
4. Ghahramani, Z. , (2001), An Introduction to Hidden Markov
Models and Bayesian Networks, J. Pattern Recognitions and
Artificial Intelligence.
5. Kalman, R. E. and Bucy, R. S. Z. , (1961), New Results on Linear
Filtering and Prediction Theory, J. Basic Engineering.