Squeezing Every Ounce of Information from An Experiment: Adaptive Design Optimization

Size: px
Start display at page:

Download "Squeezing Every Ounce of Information from An Experiment: Adaptive Design Optimization"

Transcription

1 Squeezing Every Ounce of Information from An Experiment: Adaptive Design Optimization Jay Myung Department of Psychology Ohio State University UCI Department of Cognitive Sciences Colloquium (May 21, 2014) 1

2 Wordle View of Current Research Efforts 2

3 Outline o Introduction o Adaptive Design Optimization (ADO) o Example Applications of ADO n Memory Retention Experiment n Risky Choice Experiment o Conclusions 3

4 Introduction o Experiments are fundamental to the advancement of psychological science 4

5 Introduction o Data obtained from experiments are used to fit formal models. Multinomial Processing Trees Structural Equations (SEM) Neural Network 5

6 Introduction o Often, there are many competing models to describe the same cognitive/perceptual process. o Example: Some models of memory retention (forgetting) Power Exponential Hyperbolic 6

7 Experiments to discriminate between models Flow chart of typical investigation Formulate models Experiment/ Collect data Analyze data Statistical model selection methods (e.g., AIC & BIC) 7 Model selection criteria are limited by the data they have to work with.

8 Experiments to discriminate between models Flow chart of typical investigation Formulate models Experiment/ Collect data Analyze data Collect smarter data to highlight differences between models 8

9 Collecting smarter data o The design of an experiment determines the quality of data that are collected. Design variables include: treatment levels stimulus levels number of observations 9

10 Hard-to-recruit participants 10

11 Outline o Introduction o Adaptive Design Optimization (ADO) o Example Applications of ADO n Memory Retention Experiment n Risky Choice Experiment o Conclusions 11

12 Adaptive design optimization (ADO) o Adaptively designed experiments n n Conduct the full experiment as a sequence of miniexperiments Improve the design of the next mini-experiment using knowledge gained from the previous miniexperiments Myung & Pitt (2009) Psychological Review Cavagnaro, Myung, Pitt & Kujala (2010) Neural Computation Myung, Cavagnaro & Pitt (2013) Journal of Mathematical Psychology 12

13 Adaptive design optimization (ADO) o Sequential design framework EXPERIMENTS Y obs (d 1 ) Y obs (d 2 ) Y obs (d s ) d 1 d 2 d 3 d s DESIGNS Adapt the design of the next experiment based on the results of preceding experiments 13

14 Adaptive design optimization (ADO) o Bayesian decision theoretic framework Posterior Bayesian Updating Prior Optimal Design Observed Outcome Design Optimization Experiment 14

15 Finding optimal designs o A principled approach from Bayesian decision theory n n Treat each possible design as a gamble whose payoff is determined by the outcome of an experiment carried out with that design. Compute an expected utility of each design by taking an expectation over models, parameters, and experiment outcomes. n The design with the highest expected utility is then chosen as the optimal design Optimal design: 15

16 Finding optimal designs o A principled approach from Bayesian decision theory n n Treat each possible design as a gamble whose payoff is determined by the outcome of an experiment carried out with that design. Compute an expected utility of each design by taking an expectation over models, parameters, and experiment outcomes. Value of a hypothetical experiment with design d in the case in which the true model is m, with parameters θ m, and outcome is y is observed. 16

17 Finding optimal designs o A principled approach from Bayesian decision theory n n Treat each possible design as a gamble whose payoff is determined by the outcome of an experiment carried out with that design. Compute an expected utility of each design by taking an expectation over models, parameters, and experiment outcomes. Likelihood function; E.g., binomial in f(θ m ) 17

18 Finding optimal designs o A principled approach from Bayesian decision theory n n Treat each possible design as a gamble whose payoff is determined by the outcome of an experiment carried out with that design. Compute an expected utility of each design by taking an expectation over models, parameters, and experiment outcomes. priors 18

19 Finding optimal designs o A principled approach from Bayesian decision theory n n Treat each possible design as a gamble whose payoff is determined by the outcome of an experiment carried out with that design. Compute an expected utility of each design by taking an expectation over models, parameters, and experiment outcomes. Priors can be updated between stages 19

20 Bayesian updating o Model posterior at stage S = 1,2,.. n Updated from Bayes factor calculation p0( m) p ( m) = ( m = 1,..., K) S K p ( k) BF ( y( d )) S 1 k = 1 0 ( km, ) S p ( θ ) o Parameter posterior at stage S = 1,2, n Updated using Bayes rule py ( S θm, ds) ps 1( θm) ps( θm) = ( m = 1,..., K) py ( θ, d ) p ( θ ) dθ S m S S 1 m m 20

21 Adaptive design optimization (ADO) (38) (5) EXPERIMENT y obs (d 1 ) y obs (d 2 ) y obs (d S ) p 0 (θ m ) p 0 (m) p 1 (θ m ) p 1 (m) p 2 (θ m ) p 2 (m) p S (θ m ) p S (m) d 1 d 2 d S (3s) (18s) OPTIMAL DESIGNS 21

22 Utility function o Selection of a utility function that adequately captures the goals of the experiment is an integral part of ADO. 22

23 Utility function o Selection of a utility function that adequately captures the goals of the experiment is an integral part of ADO. Take U(d) to be the mutual information of the random variables Y d and M Entropy of M Conditional entropy of M given Y and d Essentially, U(d) measures the amount of information about true model that would be provided by an experiment with design d. 23

24 Computation o Finding an optimal design requires simultaneous optimization and high-dimensional integration (Muller, Sanso & De Iorio, 2004, JASA). o Computation achieved by Sequential Monte Carlo (SMC) particle filtering algorithm with simulated annealing (Amzal, Bois, Parent & Robert, 2006, JASA). 24

25 Outline o Introduction o Adaptive Design Optimization (ADO) o Example Applications of ADO n Memory Retention Experiment n Risky Choice Experiment o Conclusions 25

26 Example Application 1 of ADO: Discriminating Models of Memory Retention 26

27 Designing a retention experiment: What Time Intervals Should be Employed? o Retention: the rate of retrieval failure over time 27

28 Designing a retention Experiment: What Time Intervals Should be Employed? o Two models of retention 28

29 Designing a retention experiment: What Time Intervals Should be Employed? Model mimicry between POW and EXP POW EXP 29

30 Designing a retention experiment: What Time Intervals Should be Employed? Model predictions for a narrow range of parameters (100 Bernoulli trials) POW EXP Recall rate Recall rate Time (seconds) 0.58<a< <b<0.70 Time (seconds) 0.88<a< <b<

31 Designing a retention experiment: What Time Intervals Should be Employed? o A good design choice can aid in discrimination seconds: Bad designs! POW EXP Recall rate Recall rate Time (seconds) Time (seconds) 31

32 Designing a retention experiment: What Time Intervals Should be Employed? o A good design choice can aid in discrimination 2-4 seconds: Good designs! POW EXP Recall rate Recall rate Time (seconds) Time (seconds) 32

33 Designing a retention experiment: What Time Intervals Should be Employed? o More realistic situations require a more principled approach to finding optimal designs. POW EXP Recall rate Recall rate Time (seconds) a~beta(2,1) b~beta(1,4) Time (seconds) a~beta(2,1) b~beta(1,80) 33

34 ADO in Action 34

35 Simulation experiment o Data generated from EXP with a=0.71 and b=

36 Simulation results 30 o Stage 1 Compute optimal design Generate data (30 Bernoulli trials) 30 P 0 (POW)=0.5 P 0 (EXP)= Time (seconds) 7 correct responses 16 seconds Time (seconds) 36

37 Simulation results 30 o Stage 1 Update model probabilities Update parameter probabilities 30 P 1 (POW)=0.65 P 1 (EXP)= Time (seconds) 7 correct responses 16 seconds Time (seconds) 37

38 Simulation results 30 o Stage 2 Compute optimal design Generate data (30 Bernoulli trials) 30 P 1 (POW)=0.65 P 1 (EXP)= Time (seconds) 0 correct responses 96.4 seconds Time (seconds) 38

39 Simulation results 30 o Stage 2 Update model probabilities Update parameter probabilities 30 P 2 (POW)=0.30 P 2 (EXP)= Time (seconds) 0 correct responses 96.4 seconds Time (seconds) 39

40 Simulation results 30 o Stage 3 Compute optimal design Generate data (30 Bernoulli trials) 30 P 2 (POW)=0.30 P 2 (EXP)= Time (seconds) 0 correct responses 96.8 seconds Time (seconds) 40

41 Simulation results o Stage 3 Update model probabilities 30 P 3 (POW)= P 3 (EXP)= Time (seconds) 0 correct responses 96.8 seconds Time (seconds) 41

42 Simulation results Bayes Factor = 20 (very strong evidence for EXP) 42

43 Simulation results Bayes Factor = 20 (very strong evidence for EXP) 43

44 Outline o Introduction o Adaptive Design Optimization (ADO) o Example Applications of ADO n Memory Retention Experiment n Risky Choice Experiment o Conclusions 44

45 Example Application 2 of ADO: Discriminating Models of Risky Choice 45

46 Risky Choice Experiment q 100 choices over the course of 60 minutes.

47 Probability Weighting Functions Empirical evidence has shown that decision makers do not weight probabilities linearly. These distortions of the probability scale affect how people choose between uncertain prospects. In Cumulative Prospect Theory (CPT), such distortions are quantified with a probability weighting function. Numerous functional forms have been proposed.

48 Probability Weighting Functions Tversky and Kahneman (1992)

49 Probability Weighting Functions Axiomatically derived functions (Prelec, 1998)

50 Probability Weighting Functions Linear in log odds (Gonzalez and Wu, 1999)

51 Probability Weighting Functions Previous attempt to discriminate probability weighting functions have yielded ambiguous results (Gonzalez & Wu, 1999; Stott, 2006)

52 Model Mimicry Given that these different functions can mimic each other so closely, does it really matter which one we use? Are there ANY situations in which these two functions would imply different choice predictions?

53 Choice Predictions of CPT Which do you prefer? Gamble A Gamble B or $0 $500 $1000 $0 $500 $1000

54 Choice Predictions of CPT Which do you prefer? Gamble A Gamble B or $0 $500 $1000 $0 $500 $1000 Weight of the probability of getting at least x i minus the weight of getting something strictly better than x i Subjective value of money

55 Choice Predictions of CPT Which do you prefer? Gamble A $0 $500 $1000 or Gamble B $0 $500 $1000 Assume WLOG that v($0)=0, v($1000) = 1, and v($500) = v where 0<v<1.

56 Choice Predictions of CPT $0 Gamble A $500 $1000 Which do you prefer? or Gamble B $500 Assuming v=0.5 and a Prelec-2 weighting function with r=0. 58 and s=1.18 So A is preferred to B $0 U(A) = > U(B) = $1000

57 Choice Predictions of CPT Gamble A $0 $500 $1000 Which do you prefer? Gamble B Assuming v=0.5 and a LinLog weighting function with r=0. 60 and s=0.65 So B is preferred to A! or U(A) = < U(B) = $0 $500 $1000

58 Choice Predictions of CPT Was this a pathological case or are there many gamble pairs with this property? One way to answer this question would be to consider the space of all possible gambles on three fixed outcomes, and search for pairs of gamble with this property. General form of a three-outcome gamble: p(x Low ) p(x Mid ) p(x High ) X Low X Mid X High

59 Designing an experiment that optimally discriminates among different forms of probability weighting function T-K Prelec-1 Prelec-2 LinLog

60 ADO-based Risky Choice Experiment q 100 choices over the course of 60 minutes. q Possible outcomes were $25, $350, $1000. q Gambles were hypothetical. Participants were paid $10. q Compared models (functional forms) based on their posterior probabilities. q Goal was to identify one form as superior (probability > 0.95)

61 Risky Choice Experiment

62 Individual results q Results for a typical participant Model probabilities across stages of the experiment MLE of each form at the conclusion of the experiment

63 Individual results q Results for another typical participant Model probabilities across stages of the experiment MLE of each form at the conclusion of the experiment

64 Individual results q Results for yet another typical participant Model probabilities across stages of the experiment MLE of each form at the conclusion of the experiment

65 Summary of results Participant ID Best Model Posterior Probability Participant ID Best Model Posterior Probability 1 LinLog Prl Prl Prl Prl EU.92 4 Prl LinLog.57 5 LinLog Prl LinLog Prl LinLog Prl Prl LinLog.80 9 EU Prl LinLog.95

66 Summary of results q Using ADO, different functional forms of the probability weighting function are discriminated decisively q One parameter functions (TK and Prl1) do not fit well at an individual participant level.. q There is striking heterogeneity in individual weighting functions.

67 Outline o Introduction o Adaptive Design Optimization (ADO) o Example Applications of ADO n Memory Retention Experiment n Risky Choice Experiment o Conclusions 67

68 Conclusions o Adaptive design optimization (ADO) is a promising new experimental tool that facilitates efficient collection of data in experiments discriminating and estimating formal models o Current and future work n n n n n n ADO for visual psychophysics ADO for cognitive neuroscience (e.g., fmri) ADO for skill learning (with AFRL) ADO for inter-temporal choice ADO with clinical patients (e.g., OCD) Hierarchical Bayes ADO 68

69 Hierarchical Bayes ADO (HADO) (Kim, Pitt, Lu, Steyvers & Myung, 2014) Posterior Predictive Joint Posterior Hierarchical Updating ADO Posterior Prior Parameter Updating Optimal Design Observed Outcome Design Optimization Experiment 69

70 Lab team and collaborators Mark Pitt Dan Cavagnaro Woojae Kim Zhong-Lin Lu (OSU) Hairong Gu (OSU) Yun Tang (OSU) Mark Steyvers (UCI) Rich Gonzalez (U Michigan) Gabe Aranovich (Stanford) Sam McClure (Stanford) 70

71 Much ADO about Nothing - William Shakespeare 71

72 Thank You! 72

Model Evaluation and Selection in Cognitive Computational Modeling

Model Evaluation and Selection in Cognitive Computational Modeling Model Evaluation and Selection in Cognitive Computational Modeling Jay Myung Department of Psychology Ohio State University, Columbus OH, USA In collaboration with Mark Pitt Mathematical Psychology Workshop

More information

Bayesian Sequential Design under Model Uncertainty using Sequential Monte Carlo

Bayesian Sequential Design under Model Uncertainty using Sequential Monte Carlo Bayesian Sequential Design under Model Uncertainty using Sequential Monte Carlo, James McGree, Tony Pettitt October 7, 2 Introduction Motivation Model choice abundant throughout literature Take into account

More information

Adaptive Design Optimization: A Mutual Information Based Approach to Model Discrimination in Cognitive Science

Adaptive Design Optimization: A Mutual Information Based Approach to Model Discrimination in Cognitive Science Adaptive Design Optimization: A Mutual Information Based Approach to Model Discrimination in Cognitive Science Daniel R. Cavagnaro Jay I. Myung Mark A. Pitt Janne V. Kujala May 26, 2009 Abstract Discriminating

More information

+ + ( + ) = Linear recurrent networks. Simpler, much more amenable to analytic treatment E.g. by choosing

+ + ( + ) = Linear recurrent networks. Simpler, much more amenable to analytic treatment E.g. by choosing Linear recurrent networks Simpler, much more amenable to analytic treatment E.g. by choosing + ( + ) = Firing rates can be negative Approximates dynamics around fixed point Approximation often reasonable

More information

CS540 Machine learning L9 Bayesian statistics

CS540 Machine learning L9 Bayesian statistics CS540 Machine learning L9 Bayesian statistics 1 Last time Naïve Bayes Beta-Bernoulli 2 Outline Bayesian concept learning Beta-Bernoulli model (review) Dirichlet-multinomial model Credible intervals 3 Bayesian

More information

9/12/17. Types of learning. Modeling data. Supervised learning: Classification. Supervised learning: Regression. Unsupervised learning: Clustering

9/12/17. Types of learning. Modeling data. Supervised learning: Classification. Supervised learning: Regression. Unsupervised learning: Clustering Types of learning Modeling data Supervised: we know input and targets Goal is to learn a model that, given input data, accurately predicts target data Unsupervised: we know the input only and want to make

More information

Readings: K&F: 16.3, 16.4, Graphical Models Carlos Guestrin Carnegie Mellon University October 6 th, 2008

Readings: K&F: 16.3, 16.4, Graphical Models Carlos Guestrin Carnegie Mellon University October 6 th, 2008 Readings: K&F: 16.3, 16.4, 17.3 Bayesian Param. Learning Bayesian Structure Learning Graphical Models 10708 Carlos Guestrin Carnegie Mellon University October 6 th, 2008 10-708 Carlos Guestrin 2006-2008

More information

Logistic Regression. Machine Learning Fall 2018

Logistic Regression. Machine Learning Fall 2018 Logistic Regression Machine Learning Fall 2018 1 Where are e? We have seen the folloing ideas Linear models Learning as loss minimization Bayesian learning criteria (MAP and MLE estimation) The Naïve Bayes

More information

Introduction to Probabilistic Machine Learning

Introduction to Probabilistic Machine Learning Introduction to Probabilistic Machine Learning Piyush Rai Dept. of CSE, IIT Kanpur (Mini-course 1) Nov 03, 2015 Piyush Rai (IIT Kanpur) Introduction to Probabilistic Machine Learning 1 Machine Learning

More information

Hierarchical Models & Bayesian Model Selection

Hierarchical Models & Bayesian Model Selection Hierarchical Models & Bayesian Model Selection Geoffrey Roeder Departments of Computer Science and Statistics University of British Columbia Jan. 20, 2016 Contact information Please report any typos or

More information

Introduction to Bayesian Statistics

Introduction to Bayesian Statistics Bayesian Parameter Estimation Introduction to Bayesian Statistics Harvey Thornburg Center for Computer Research in Music and Acoustics (CCRMA) Department of Music, Stanford University Stanford, California

More information

Bayesian Networks: Construction, Inference, Learning and Causal Interpretation. Volker Tresp Summer 2016

Bayesian Networks: Construction, Inference, Learning and Causal Interpretation. Volker Tresp Summer 2016 Bayesian Networks: Construction, Inference, Learning and Causal Interpretation Volker Tresp Summer 2016 1 Introduction So far we were mostly concerned with supervised learning: we predicted one or several

More information

Lecture 4. Generative Models for Discrete Data - Part 3. Luigi Freda. ALCOR Lab DIAG University of Rome La Sapienza.

Lecture 4. Generative Models for Discrete Data - Part 3. Luigi Freda. ALCOR Lab DIAG University of Rome La Sapienza. Lecture 4 Generative Models for Discrete Data - Part 3 Luigi Freda ALCOR Lab DIAG University of Rome La Sapienza October 6, 2017 Luigi Freda ( La Sapienza University) Lecture 4 October 6, 2017 1 / 46 Outline

More information

Harrison B. Prosper. Bari Lectures

Harrison B. Prosper. Bari Lectures Harrison B. Prosper Florida State University Bari Lectures 30, 31 May, 1 June 2016 Lectures on Multivariate Methods Harrison B. Prosper Bari, 2016 1 h Lecture 1 h Introduction h Classification h Grid Searches

More information

Last Time. Today. Bayesian Learning. The Distributions We Love. CSE 446 Gaussian Naïve Bayes & Logistic Regression

Last Time. Today. Bayesian Learning. The Distributions We Love. CSE 446 Gaussian Naïve Bayes & Logistic Regression CSE 446 Gaussian Naïve Bayes & Logistic Regression Winter 22 Dan Weld Learning Gaussians Naïve Bayes Last Time Gaussians Naïve Bayes Logistic Regression Today Some slides from Carlos Guestrin, Luke Zettlemoyer

More information

CS 6140: Machine Learning Spring What We Learned Last Week. Survey 2/26/16. VS. Model

CS 6140: Machine Learning Spring What We Learned Last Week. Survey 2/26/16. VS. Model Logis@cs CS 6140: Machine Learning Spring 2016 Instructor: Lu Wang College of Computer and Informa@on Science Northeastern University Webpage: www.ccs.neu.edu/home/luwang Email: luwang@ccs.neu.edu Assignment

More information

Today s Outline. Recap: MDPs. Bellman Equations. Q-Value Iteration. Bellman Backup 5/7/2012. CSE 473: Artificial Intelligence Reinforcement Learning

Today s Outline. Recap: MDPs. Bellman Equations. Q-Value Iteration. Bellman Backup 5/7/2012. CSE 473: Artificial Intelligence Reinforcement Learning CSE 473: Artificial Intelligence Reinforcement Learning Dan Weld Today s Outline Reinforcement Learning Q-value iteration Q-learning Exploration / exploitation Linear function approximation Many slides

More information

CPSC 540: Machine Learning

CPSC 540: Machine Learning CPSC 540: Machine Learning Empirical Bayes, Hierarchical Bayes Mark Schmidt University of British Columbia Winter 2017 Admin Assignment 5: Due April 10. Project description on Piazza. Final details coming

More information

CS 6140: Machine Learning Spring 2016

CS 6140: Machine Learning Spring 2016 CS 6140: Machine Learning Spring 2016 Instructor: Lu Wang College of Computer and Informa?on Science Northeastern University Webpage: www.ccs.neu.edu/home/luwang Email: luwang@ccs.neu.edu Logis?cs Assignment

More information

Preliminary Statistics Lecture 2: Probability Theory (Outline) prelimsoas.webs.com

Preliminary Statistics Lecture 2: Probability Theory (Outline) prelimsoas.webs.com 1 School of Oriental and African Studies September 2015 Department of Economics Preliminary Statistics Lecture 2: Probability Theory (Outline) prelimsoas.webs.com Gujarati D. Basic Econometrics, Appendix

More information

Click Prediction and Preference Ranking of RSS Feeds

Click Prediction and Preference Ranking of RSS Feeds Click Prediction and Preference Ranking of RSS Feeds 1 Introduction December 11, 2009 Steven Wu RSS (Really Simple Syndication) is a family of data formats used to publish frequently updated works. RSS

More information

Basic Probabilistic Reasoning SEG

Basic Probabilistic Reasoning SEG Basic Probabilistic Reasoning SEG 7450 1 Introduction Reasoning under uncertainty using probability theory Dealing with uncertainty is one of the main advantages of an expert system over a simple decision

More information

Hypothesis Testing. Part I. James J. Heckman University of Chicago. Econ 312 This draft, April 20, 2006

Hypothesis Testing. Part I. James J. Heckman University of Chicago. Econ 312 This draft, April 20, 2006 Hypothesis Testing Part I James J. Heckman University of Chicago Econ 312 This draft, April 20, 2006 1 1 A Brief Review of Hypothesis Testing and Its Uses values and pure significance tests (R.A. Fisher)

More information

Sequential Monte Carlo Algorithms for Bayesian Sequential Design

Sequential Monte Carlo Algorithms for Bayesian Sequential Design Sequential Monte Carlo Algorithms for Bayesian Sequential Design Dr Queensland University of Technology c.drovandi@qut.edu.au Collaborators: James McGree, Tony Pettitt, Gentry White Acknowledgements: Australian

More information

Predictive Processing in Planning:

Predictive Processing in Planning: Predictive Processing in Planning: Choice Behavior as Active Bayesian Inference Philipp Schwartenbeck Wellcome Trust Centre for Human Neuroimaging, UCL The Promise of Predictive Processing: A Critical

More information

Class 26: review for final exam 18.05, Spring 2014

Class 26: review for final exam 18.05, Spring 2014 Probability Class 26: review for final eam 8.05, Spring 204 Counting Sets Inclusion-eclusion principle Rule of product (multiplication rule) Permutation and combinations Basics Outcome, sample space, event

More information

Bayesian statistics. DS GA 1002 Statistical and Mathematical Models. Carlos Fernandez-Granda

Bayesian statistics. DS GA 1002 Statistical and Mathematical Models.   Carlos Fernandez-Granda Bayesian statistics DS GA 1002 Statistical and Mathematical Models http://www.cims.nyu.edu/~cfgranda/pages/dsga1002_fall15 Carlos Fernandez-Granda Frequentist vs Bayesian statistics In frequentist statistics

More information

Structure learning in human causal induction

Structure learning in human causal induction Structure learning in human causal induction Joshua B. Tenenbaum & Thomas L. Griffiths Department of Psychology Stanford University, Stanford, CA 94305 jbt,gruffydd @psych.stanford.edu Abstract We use

More information

Some slides from Carlos Guestrin, Luke Zettlemoyer & K Gajos 2

Some slides from Carlos Guestrin, Luke Zettlemoyer & K Gajos 2 Logistics CSE 446: Point Estimation Winter 2012 PS2 out shortly Dan Weld Some slides from Carlos Guestrin, Luke Zettlemoyer & K Gajos 2 Last Time Random variables, distributions Marginal, joint & conditional

More information

Logistic Regression. Seungjin Choi

Logistic Regression. Seungjin Choi Logistic Regression Seungjin Choi Department of Computer Science and Engineering Pohang University of Science and Technology 77 Cheongam-ro, Nam-gu, Pohang 37673, Korea seungjin@postech.ac.kr http://mlg.postech.ac.kr/

More information

Time Series and Dynamic Models

Time Series and Dynamic Models Time Series and Dynamic Models Section 1 Intro to Bayesian Inference Carlos M. Carvalho The University of Texas at Austin 1 Outline 1 1. Foundations of Bayesian Statistics 2. Bayesian Estimation 3. The

More information

Machine Learning

Machine Learning Machine Learning 10-601 Tom M. Mitchell Machine Learning Department Carnegie Mellon University August 30, 2017 Today: Decision trees Overfitting The Big Picture Coming soon Probabilistic learning MLE,

More information

Grundlagen der Künstlichen Intelligenz

Grundlagen der Künstlichen Intelligenz Grundlagen der Künstlichen Intelligenz Uncertainty & Probabilities & Bandits Daniel Hennes 16.11.2017 (WS 2017/18) University Stuttgart - IPVS - Machine Learning & Robotics 1 Today Uncertainty Probability

More information

Sequential Decisions

Sequential Decisions Sequential Decisions A Basic Theorem of (Bayesian) Expected Utility Theory: If you can postpone a terminal decision in order to observe, cost free, an experiment whose outcome might change your terminal

More information

Journal of Mathematical Psychology

Journal of Mathematical Psychology Journal of Mathematical Psychology 56 () 8 85 Contents lists available at SciVerse ScienceDirect Journal of Mathematical Psychology journal homepage: www.elsevier.com/locate/jmp Notes and comment Bayesian

More information

Algorithm-Independent Learning Issues

Algorithm-Independent Learning Issues Algorithm-Independent Learning Issues Selim Aksoy Department of Computer Engineering Bilkent University saksoy@cs.bilkent.edu.tr CS 551, Spring 2007 c 2007, Selim Aksoy Introduction We have seen many learning

More information

COMP90051 Statistical Machine Learning

COMP90051 Statistical Machine Learning COMP90051 Statistical Machine Learning Semester 2, 2017 Lecturer: Trevor Cohn 2. Statistical Schools Adapted from slides by Ben Rubinstein Statistical Schools of Thought Remainder of lecture is to provide

More information

ECE521 Lecture7. Logistic Regression

ECE521 Lecture7. Logistic Regression ECE521 Lecture7 Logistic Regression Outline Review of decision theory Logistic regression A single neuron Multi-class classification 2 Outline Decision theory is conceptually easy and computationally hard

More information

DS-GA 1002 Lecture notes 11 Fall Bayesian statistics

DS-GA 1002 Lecture notes 11 Fall Bayesian statistics DS-GA 100 Lecture notes 11 Fall 016 Bayesian statistics In the frequentist paradigm we model the data as realizations from a distribution that depends on deterministic parameters. In contrast, in Bayesian

More information

Lecture 2: From Linear Regression to Kalman Filter and Beyond

Lecture 2: From Linear Regression to Kalman Filter and Beyond Lecture 2: From Linear Regression to Kalman Filter and Beyond Department of Biomedical Engineering and Computational Science Aalto University January 26, 2012 Contents 1 Batch and Recursive Estimation

More information

Chapter 4: Modelling

Chapter 4: Modelling Chapter 4: Modelling Exchangeability and Invariance Markus Harva 17.10. / Reading Circle on Bayesian Theory Outline 1 Introduction 2 Models via exchangeability 3 Models via invariance 4 Exercise Statistical

More information

Bayesian Model Diagnostics and Checking

Bayesian Model Diagnostics and Checking Earvin Balderama Quantitative Ecology Lab Department of Forestry and Environmental Resources North Carolina State University April 12, 2013 1 / 34 Introduction MCMCMC 2 / 34 Introduction MCMCMC Steps in

More information

Uncertainty, precision, prediction errors and their relevance to computational psychiatry

Uncertainty, precision, prediction errors and their relevance to computational psychiatry Uncertainty, precision, prediction errors and their relevance to computational psychiatry Christoph Mathys Wellcome Trust Centre for Neuroimaging at UCL, London, UK Max Planck UCL Centre for Computational

More information

Learning with Probabilities

Learning with Probabilities Learning with Probabilities CS194-10 Fall 2011 Lecture 15 CS194-10 Fall 2011 Lecture 15 1 Outline Bayesian learning eliminates arbitrary loss functions and regularizers facilitates incorporation of prior

More information

Common ratio using delay

Common ratio using delay Theory Dec. (2010) 68:149 158 DOI 10.1007/s18-008-9130-2 Common ratio using delay Manel Baucells Franz H. Heukamp Published online: 4 January 2009 Springer Science+Business Media, LLC. 2009 Abstract We

More information

Bayesian model selection: methodology, computation and applications

Bayesian model selection: methodology, computation and applications Bayesian model selection: methodology, computation and applications David Nott Department of Statistics and Applied Probability National University of Singapore Statistical Genomics Summer School Program

More information

Probabilistic Graphical Models

Probabilistic Graphical Models Probabilistic Graphical Models Lecture 11 CRFs, Exponential Family CS/CNS/EE 155 Andreas Krause Announcements Homework 2 due today Project milestones due next Monday (Nov 9) About half the work should

More information

Bayesian Networks: Construction, Inference, Learning and Causal Interpretation. Volker Tresp Summer 2014

Bayesian Networks: Construction, Inference, Learning and Causal Interpretation. Volker Tresp Summer 2014 Bayesian Networks: Construction, Inference, Learning and Causal Interpretation Volker Tresp Summer 2014 1 Introduction So far we were mostly concerned with supervised learning: we predicted one or several

More information

Multi-Attribute Bayesian Optimization under Utility Uncertainty

Multi-Attribute Bayesian Optimization under Utility Uncertainty Multi-Attribute Bayesian Optimization under Utility Uncertainty Raul Astudillo Cornell University Ithaca, NY 14853 ra598@cornell.edu Peter I. Frazier Cornell University Ithaca, NY 14853 pf98@cornell.edu

More information

Choosing among models

Choosing among models Eco 515 Fall 2014 Chris Sims Choosing among models September 18, 2014 c 2014 by Christopher A. Sims. This document is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported

More information

Probability and Estimation. Alan Moses

Probability and Estimation. Alan Moses Probability and Estimation Alan Moses Random variables and probability A random variable is like a variable in algebra (e.g., y=e x ), but where at least part of the variability is taken to be stochastic.

More information

Bayesian Networks Inference with Probabilistic Graphical Models

Bayesian Networks Inference with Probabilistic Graphical Models 4190.408 2016-Spring Bayesian Networks Inference with Probabilistic Graphical Models Byoung-Tak Zhang intelligence Lab Seoul National University 4190.408 Artificial (2016-Spring) 1 Machine Learning? Learning

More information

Estimation of reliability parameters from Experimental data (Parte 2) Prof. Enrico Zio

Estimation of reliability parameters from Experimental data (Parte 2) Prof. Enrico Zio Estimation of reliability parameters from Experimental data (Parte 2) This lecture Life test (t 1,t 2,...,t n ) Estimate θ of f T t θ For example: λ of f T (t)= λe - λt Classical approach (frequentist

More information

Bayesian Social Learning with Random Decision Making in Sequential Systems

Bayesian Social Learning with Random Decision Making in Sequential Systems Bayesian Social Learning with Random Decision Making in Sequential Systems Yunlong Wang supervised by Petar M. Djurić Department of Electrical and Computer Engineering Stony Brook University Stony Brook,

More information

Bayesian Methods for Machine Learning

Bayesian Methods for Machine Learning Bayesian Methods for Machine Learning CS 584: Big Data Analytics Material adapted from Radford Neal s tutorial (http://ftp.cs.utoronto.ca/pub/radford/bayes-tut.pdf), Zoubin Ghahramni (http://hunch.net/~coms-4771/zoubin_ghahramani_bayesian_learning.pdf),

More information

Interim Monitoring of Clinical Trials: Decision Theory, Dynamic Programming. and Optimal Stopping

Interim Monitoring of Clinical Trials: Decision Theory, Dynamic Programming. and Optimal Stopping Interim Monitoring of Clinical Trials: Decision Theory, Dynamic Programming and Optimal Stopping Christopher Jennison Department of Mathematical Sciences, University of Bath, UK http://people.bath.ac.uk/mascj

More information

Chapter 6. Bayesian Adaptive Estimation: A Theoretical Review

Chapter 6. Bayesian Adaptive Estimation: A Theoretical Review Chapter 6 Bayesian Adaptive Estimation: A Theoretical Review Janne V. Kujala Department of Mathematical Information Technology, University of Jyväskylä, P.O.Box 35 (Agora), FI-4004 University of Jyväskylä,

More information

Computational Cognitive Science

Computational Cognitive Science Computational Cognitive Science Lecture 9: Bayesian Estimation Chris Lucas (Slides adapted from Frank Keller s) School of Informatics University of Edinburgh clucas2@inf.ed.ac.uk 17 October, 2017 1 / 28

More information

Clustering and Gaussian Mixture Models

Clustering and Gaussian Mixture Models Clustering and Gaussian Mixture Models Piyush Rai IIT Kanpur Probabilistic Machine Learning (CS772A) Jan 25, 2016 Probabilistic Machine Learning (CS772A) Clustering and Gaussian Mixture Models 1 Recap

More information

Hybrid Bayesian-frequentist approaches for small sample trial design: examples and discussion on concepts.

Hybrid Bayesian-frequentist approaches for small sample trial design: examples and discussion on concepts. Hybrid Bayesian-frequentist approaches for small sample trial design: examples and discussion on concepts. Stavros Nikolakopoulos Kit Roes UMC Utrecht Outline Comfortable or not with hybrid Bayesian-frequentist

More information

David Giles Bayesian Econometrics

David Giles Bayesian Econometrics David Giles Bayesian Econometrics 1. General Background 2. Constructing Prior Distributions 3. Properties of Bayes Estimators and Tests 4. Bayesian Analysis of the Multiple Regression Model 5. Bayesian

More information

Algebraic Geometry and Model Selection

Algebraic Geometry and Model Selection Algebraic Geometry and Model Selection American Institute of Mathematics 2011/Dec/12-16 I would like to thank Prof. Russell Steele, Prof. Bernd Sturmfels, and all participants. Thank you very much. Sumio

More information

Introduction to Bayesian Inference

Introduction to Bayesian Inference Introduction to Bayesian Inference p. 1/2 Introduction to Bayesian Inference September 15th, 2010 Reading: Hoff Chapter 1-2 Introduction to Bayesian Inference p. 2/2 Probability: Measurement of Uncertainty

More information

Machine Learning

Machine Learning Machine Learning 10-701 Tom M. Mitchell Machine Learning Department Carnegie Mellon University January 13, 2011 Today: The Big Picture Overfitting Review: probability Readings: Decision trees, overfiting

More information

Generative Models for Discrete Data

Generative Models for Discrete Data Generative Models for Discrete Data ddebarr@uw.edu 2016-04-21 Agenda Bayesian Concept Learning Beta-Binomial Model Dirichlet-Multinomial Model Naïve Bayes Classifiers Bayesian Concept Learning Numbers

More information

Machine Learning. Theory of Classification and Nonparametric Classifier. Lecture 2, January 16, What is theoretically the best classifier

Machine Learning. Theory of Classification and Nonparametric Classifier. Lecture 2, January 16, What is theoretically the best classifier Machine Learning 10-701/15 701/15-781, 781, Spring 2008 Theory of Classification and Nonparametric Classifier Eric Xing Lecture 2, January 16, 2006 Reading: Chap. 2,5 CB and handouts Outline What is theoretically

More information

Miscellany : Long Run Behavior of Bayesian Methods; Bayesian Experimental Design (Lecture 4)

Miscellany : Long Run Behavior of Bayesian Methods; Bayesian Experimental Design (Lecture 4) Miscellany : Long Run Behavior of Bayesian Methods; Bayesian Experimental Design (Lecture 4) Tom Loredo Dept. of Astronomy, Cornell University http://www.astro.cornell.edu/staff/loredo/bayes/ Bayesian

More information

Inference for a Population Proportion

Inference for a Population Proportion Al Nosedal. University of Toronto. November 11, 2015 Statistical inference is drawing conclusions about an entire population based on data in a sample drawn from that population. From both frequentist

More information

7. Estimation and hypothesis testing. Objective. Recommended reading

7. Estimation and hypothesis testing. Objective. Recommended reading 7. Estimation and hypothesis testing Objective In this chapter, we show how the election of estimators can be represented as a decision problem. Secondly, we consider the problem of hypothesis testing

More information

Uniform Sources of Uncertainty for Subjective Probabilities and

Uniform Sources of Uncertainty for Subjective Probabilities and Uniform Sources of Uncertainty for Subjective Probabilities and Ambiguity Mohammed Abdellaoui (joint with Aurélien Baillon and Peter Wakker) 1 Informal Central in this work will be the recent finding of

More information

Fundamentals. CS 281A: Statistical Learning Theory. Yangqing Jia. August, Based on tutorial slides by Lester Mackey and Ariel Kleiner

Fundamentals. CS 281A: Statistical Learning Theory. Yangqing Jia. August, Based on tutorial slides by Lester Mackey and Ariel Kleiner Fundamentals CS 281A: Statistical Learning Theory Yangqing Jia Based on tutorial slides by Lester Mackey and Ariel Kleiner August, 2011 Outline 1 Probability 2 Statistics 3 Linear Algebra 4 Optimization

More information

Bayesian Networks BY: MOHAMAD ALSABBAGH

Bayesian Networks BY: MOHAMAD ALSABBAGH Bayesian Networks BY: MOHAMAD ALSABBAGH Outlines Introduction Bayes Rule Bayesian Networks (BN) Representation Size of a Bayesian Network Inference via BN BN Learning Dynamic BN Introduction Conditional

More information

Optimising Group Sequential Designs. Decision Theory, Dynamic Programming. and Optimal Stopping

Optimising Group Sequential Designs. Decision Theory, Dynamic Programming. and Optimal Stopping : Decision Theory, Dynamic Programming and Optimal Stopping Christopher Jennison Department of Mathematical Sciences, University of Bath, UK http://people.bath.ac.uk/mascj InSPiRe Conference on Methodology

More information

Towards a Theory of Decision Making without Paradoxes

Towards a Theory of Decision Making without Paradoxes Towards a Theory of Decision Making without Paradoxes Roman V. Belavkin (R.Belavkin@mdx.ac.uk) School of Computing Science, Middlesex University London NW4 4BT, United Kingdom Abstract Human subjects often

More information

Lecture : Probabilistic Machine Learning

Lecture : Probabilistic Machine Learning Lecture : Probabilistic Machine Learning Riashat Islam Reasoning and Learning Lab McGill University September 11, 2018 ML : Many Methods with Many Links Modelling Views of Machine Learning Machine Learning

More information

Statistics for the LHC Lecture 1: Introduction

Statistics for the LHC Lecture 1: Introduction Statistics for the LHC Lecture 1: Introduction Academic Training Lectures CERN, 14 17 June, 2010 indico.cern.ch/conferencedisplay.py?confid=77830 Glen Cowan Physics Department Royal Holloway, University

More information

Classification and Regression Trees

Classification and Regression Trees Classification and Regression Trees Ryan P Adams So far, we have primarily examined linear classifiers and regressors, and considered several different ways to train them When we ve found the linearity

More information

A Bayesian model for event-based trust

A Bayesian model for event-based trust A Bayesian model for event-based trust Elements of a foundation for computational trust Vladimiro Sassone ECS, University of Southampton joint work K. Krukow and M. Nielsen Oxford, 9 March 2007 V. Sassone

More information

Clinical Trials. Olli Saarela. September 18, Dalla Lana School of Public Health University of Toronto.

Clinical Trials. Olli Saarela. September 18, Dalla Lana School of Public Health University of Toronto. Introduction to Dalla Lana School of Public Health University of Toronto olli.saarela@utoronto.ca September 18, 2014 38-1 : a review 38-2 Evidence Ideal: to advance the knowledge-base of clinical medicine,

More information

Statistical Rock Physics

Statistical Rock Physics Statistical - Introduction Book review 3.1-3.3 Min Sun March. 13, 2009 Outline. What is Statistical. Why we need Statistical. How Statistical works Statistical Rock physics Information theory Statistics

More information

CS 630 Basic Probability and Information Theory. Tim Campbell

CS 630 Basic Probability and Information Theory. Tim Campbell CS 630 Basic Probability and Information Theory Tim Campbell 21 January 2003 Probability Theory Probability Theory is the study of how best to predict outcomes of events. An experiment (or trial or event)

More information

Bayesian Inference. Will Penny. 24th February Bayesian Inference. Will Penny. Bayesian Inference. References

Bayesian Inference. Will Penny. 24th February Bayesian Inference. Will Penny. Bayesian Inference. References 24th February 2011 Given probabilities p(a), p(b), and the joint probability p(a, B), we can write the conditional probabilities p(b A) = p(a B) = p(a, B) p(a) p(a, B) p(b) Eliminating p(a, B) gives p(b

More information

Supplementary Material: A Robust Approach to Sequential Information Theoretic Planning

Supplementary Material: A Robust Approach to Sequential Information Theoretic Planning Supplementar aterial: A Robust Approach to Sequential Information Theoretic Planning Sue Zheng Jason Pacheco John W. Fisher, III. Proofs of Estimator Properties.. Proof of Prop. Here we show how the bias

More information

A.I. in health informatics lecture 2 clinical reasoning & probabilistic inference, I. kevin small & byron wallace

A.I. in health informatics lecture 2 clinical reasoning & probabilistic inference, I. kevin small & byron wallace A.I. in health informatics lecture 2 clinical reasoning & probabilistic inference, I kevin small & byron wallace today a review of probability random variables, maximum likelihood, etc. crucial for clinical

More information

Lectures on Statistical Data Analysis

Lectures on Statistical Data Analysis Lectures on Statistical Data Analysis London Postgraduate Lectures on Particle Physics; University of London MSci course PH4515 Glen Cowan Physics Department Royal Holloway, University of London g.cowan@rhul.ac.uk

More information

Bayesian Statistics for Personalized Medicine. David Yang

Bayesian Statistics for Personalized Medicine. David Yang Bayesian Statistics for Personalized Medicine David Yang Outline Why Bayesian Statistics for Personalized Medicine? A Network-based Bayesian Strategy for Genomic Biomarker Discovery Part One Why Bayesian

More information

PROBABILITY AND INFORMATION THEORY. Dr. Gjergji Kasneci Introduction to Information Retrieval WS

PROBABILITY AND INFORMATION THEORY. Dr. Gjergji Kasneci Introduction to Information Retrieval WS PROBABILITY AND INFORMATION THEORY Dr. Gjergji Kasneci Introduction to Information Retrieval WS 2012-13 1 Outline Intro Basics of probability and information theory Probability space Rules of probability

More information

The Knowledge Gradient for Sequential Decision Making with Stochastic Binary Feedbacks

The Knowledge Gradient for Sequential Decision Making with Stochastic Binary Feedbacks The Knowledge Gradient for Sequential Decision Making with Stochastic Binary Feedbacks Yingfei Wang, Chu Wang and Warren B. Powell Princeton University Yingfei Wang Optimal Learning Methods June 22, 2016

More information

Probabilistic Graphical Models

Probabilistic Graphical Models Probabilistic Graphical Models Lecture 5 Bayesian Learning of Bayesian Networks CS/CNS/EE 155 Andreas Krause Announcements Recitations: Every Tuesday 4-5:30 in 243 Annenberg Homework 1 out. Due in class

More information

Related Concepts: Lecture 9 SEM, Statistical Modeling, AI, and Data Mining. I. Terminology of SEM

Related Concepts: Lecture 9 SEM, Statistical Modeling, AI, and Data Mining. I. Terminology of SEM Lecture 9 SEM, Statistical Modeling, AI, and Data Mining I. Terminology of SEM Related Concepts: Causal Modeling Path Analysis Structural Equation Modeling Latent variables (Factors measurable, but thru

More information

Answers and expectations

Answers and expectations Answers and expectations For a function f(x) and distribution P(x), the expectation of f with respect to P is The expectation is the average of f, when x is drawn from the probability distribution P E

More information

ST440/540: Applied Bayesian Statistics. (9) Model selection and goodness-of-fit checks

ST440/540: Applied Bayesian Statistics. (9) Model selection and goodness-of-fit checks (9) Model selection and goodness-of-fit checks Objectives In this module we will study methods for model comparisons and checking for model adequacy For model comparisons there are a finite number of candidate

More information

Bayesian model comparison and distinguishability

Bayesian model comparison and distinguishability Bayesian model comparison and distinguishability Julien Diard (julien.diard@upmf-grenoble.fr) Laboratoire de Psychologie et NeuroCognition CNRS-UPMF Grenoble, France Abstract This paper focuses on Bayesian

More information

Value of Information Analysis with Structural Reliability Methods

Value of Information Analysis with Structural Reliability Methods Accepted for publication in Structural Safety, special issue in the honor of Prof. Wilson Tang August 2013 Value of Information Analysis with Structural Reliability Methods Daniel Straub Engineering Risk

More information

Introduction to Machine Learning

Introduction to Machine Learning Introduction to Machine Learning Logistic Regression Varun Chandola Computer Science & Engineering State University of New York at Buffalo Buffalo, NY, USA chandola@buffalo.edu Chandola@UB CSE 474/574

More information

Introduction to Machine Learning

Introduction to Machine Learning Introduction to Machine Learning Introduction to Probabilistic Methods Varun Chandola Computer Science & Engineering State University of New York at Buffalo Buffalo, NY, USA chandola@buffalo.edu Chandola@UB

More information

COMP90051 Statistical Machine Learning

COMP90051 Statistical Machine Learning COMP90051 Statistical Machine Learning Semester 2, 2017 Lecturer: Trevor Cohn 17. Bayesian inference; Bayesian regression Training == optimisation (?) Stages of learning & inference: Formulate model Regression

More information

Lecture 2: From Linear Regression to Kalman Filter and Beyond

Lecture 2: From Linear Regression to Kalman Filter and Beyond Lecture 2: From Linear Regression to Kalman Filter and Beyond January 18, 2017 Contents 1 Batch and Recursive Estimation 2 Towards Bayesian Filtering 3 Kalman Filter and Bayesian Filtering and Smoothing

More information

Intro to Probability. Andrei Barbu

Intro to Probability. Andrei Barbu Intro to Probability Andrei Barbu Some problems Some problems A means to capture uncertainty Some problems A means to capture uncertainty You have data from two sources, are they different? Some problems

More information

Bayesian inference. Fredrik Ronquist and Peter Beerli. October 3, 2007

Bayesian inference. Fredrik Ronquist and Peter Beerli. October 3, 2007 Bayesian inference Fredrik Ronquist and Peter Beerli October 3, 2007 1 Introduction The last few decades has seen a growing interest in Bayesian inference, an alternative approach to statistical inference.

More information