Learning Bounds for Importance Weighting

Size: px
Start display at page:

Download "Learning Bounds for Importance Weighting"

Transcription

1 Learning Bounds for Importance Weighting Corinna Cortes Google Research Yishay Mansour Tel-Aviv University Mehryar Mohri Courant & Google

2 Motivation used in variety of contets: sample bias correction (e.g., Dudík et al., 2006; Zadrozny et al., 2003; Huang et al., 2006; Sugiyama et al., 2008). domain adaptation. active learning (Beygelzimer et al., 2009). analysis of boosting (Dasgupta and Long, 2003). Guarantees? when is importance weighting successful? are there better reweighting techniques than the straightforward standard approach? 2

3 Setting Input space X, output space Y. Loss function L: Y Y [0, 1]. Source distribution Q. Target distribution P. Training sample S of size m drawn according to Q. Hypothesis set H. Fied target labeling function f : X Y. Notation: for any X and, h H L h ()=L(h(),f()). 3

4 Importance Weighting Emphasize loss of training point by w() : empirical loss: generalization loss: R w (h)= 1 m m i=1 w( i) L h ( i ). R(h)= E P [L h ()]. Weight assumed known: w()=p ()/Q(). In practice, estimated weight w() : effect of this error specifically analyzed by (Cortes et al., 2008). Different scenarios: importance weighting/sampling. imp. weighting: finite sample of size m some Q. imp. sampling: unlimited sampling, can choose Q. 4

5 Does Importance Weighting Work? Q Q P P Training set size Error Ratio = 0.3 Q P Training set size Error Ratio = 0.75 Q P Hypothesis class: hyperplanes tangent to the unit circle. Best hypothesis chosen by empirical risk minimization.

6 Preliminaries. Outline Learning bounds for bounded importance weights. Learning bounds for unbounded importance weights (the most common case). Alternative reweighting techniques. 6

7 Rényi Divergences Definition: for α 0, D 1 D α D α (P Q) = 1 α 1 log 2 D α 0 Notation: for all α 0. coincides with the relative entropy (KL div.). non-decreasing function of α. d α (P Q) =2 D α(p Q) = P () α 1 P (). Q() P α () 1 α 1 Q α 1 () (Rényi, 1960). 7

8 Properties of w : Properties of : Properties E[w]=1 E[w 2 ]=d 2 (P Q) σ 2 (w)=d 2 (P Q) 1. wl h E Q [ R w (h)]= E Q [w()l h ()]= E Q [w2 () L 2 h()] = P () Q() L h() Q()= 2 P () Q() L 2 Q() h() = P ()L h ()=R(h). P () 1 α P () Q() P () α 1 α L 2 h () P () (Hölder s inequality) P () Q() = d α1 (P Q) α 1 α P () L h ()L α1 α 1 h () α 1 P () L 2α α α 1 h () α 1 α dα1 (P Q) R(h) 1 1 α. 8

9 Learning Bound - Bounded Case Assumption (bounded case): M =sup w()=sup Theorem: let H be a finite hypothesis set. Then, for any δ>0, with probability at least 1 δ, for any h H, R(h) R w (h) 2M(log H log1 δ ) 3m P () Q() =d (P Q)<. similar result for infinite hypothesis sets. note the role of Rényi divergence. 2d 2 (P Q)(log H log 1 δ ) m. 9

10 Lower Bound - Bounded Case Theorem: assume that M< and σ 2 (w)/m 2 1/m. Assume that H contains a hypothesis h 0 such that L h0 ()=1 for all. Then, there eists an absolute constant c=2/41 2 such that Pr sup h H R(h) R w (h) d2 (P Q) 1 c>0. 4m result based on proof of general lower bound theorem for maimal variance. 10

11 Assumption some natural cases. Unbounded Case Eamples: Gaussian distributions. P () = 1 2πσP ep does not hold, even in even for σ P =σ Q and µ=µ, d (P Q)=. but, for σ Q > 2 2 σ P slide 2), of w is bounded. d (P Q)< ( µ)2 2σ 2 P d 2 (P Q)< Q() = 1 2πσQ ep ( µ ) 2 2σ 2 Q (e.g., eample on the right,, thus the second-moment. 11

12 Learning Bound - Unbounded Case Theorem: let H such that is finite. Assume that for all. Then, for any least 1 δ, for all h H, Pdim({L h (): h H})=p d 2 (P Q)< w()=0 δ>0 and, with probability at R(h) R w (h)2 5/4 d 2 (P Q) 3 p log 2me 8 p log 4 δ. m holds even for unbounded weights. based on new proof of general learning bound theorem for unbounded losses with bounded second-moment (Vapnik s proof is incorrect!). 12

13 Proof Ideas 13

14 Alternative Weighting Techniques Arbitrary u: X R with u>0 and for any h H, R u (h)= 1 m m i=1 u( i )L h ( i ). Theorem: let H such that Pdim({L h (): h H})=p is finite. Assume that 0<E[u 2 ()]<. Then, for Q any δ>0, with probability at least 1 δ, for all h H, R(h) R u (h) E Q [w() u()]lh () 2 5/4 ma EQ [u 2 ()L 2 h ()], E bq [u 2 ()L 2 h ()] 38 p log 2me p log 4 δ. m 14

15 Alternative Weighting Techniques Trade-off between bias term and second moment ma E Q [(w() u())l h ()] E Q [u 2 ()L 2 h ()], E bq [u 2 ()L 2 h ()]. Using upper bound independent of optimization problem with γ>0 min u U E Q H w() u() γ E Q [u 2 ], a trade-off parameter. leads to the 15

16 Alternative Reweighting - Eample (1) (2) (3) (4) The variance is reduced in (3) by replacing the average weight per quantile. w with 16

17 Conclusion and Open Questions Learning guarantees for importance weighting, including unbounded case (most common). Analysis of cases where importance weighting can succeed. Critical role of Rényi divergence of the distributions. Preliminary eploration of other reweighting techniques. Estimation of Rényi divergence from finite samples. 17

learning bounds for importance weighting Tamas Madarasz & Michael Rabadi April 15, 2015

learning bounds for importance weighting Tamas Madarasz & Michael Rabadi April 15, 2015 learning bounds for importance weighting Tamas Madarasz & Michael Rabadi April 15, 2015 Introduction Often, training distribution does not match testing distribution Want to utilize information about test

More information

Sample Selection Bias Correction

Sample Selection Bias Correction Sample Selection Bias Correction Afshin Rostamizadeh Joint work with: Corinna Cortes, Mehryar Mohri & Michael Riley Courant Institute & Google Research Motivation Critical Assumption: Samples for training

More information

Learning with Imperfect Data

Learning with Imperfect Data Mehryar Mohri Courant Institute and Google mohri@cims.nyu.edu Joint work with: Yishay Mansour (Tel-Aviv & Google) and Afshin Rostamizadeh (Courant Institute). Standard Learning Assumptions IID assumption.

More information

Domain Adaptation for Regression

Domain Adaptation for Regression Domain Adaptation for Regression Corinna Cortes Google Research corinna@google.com Mehryar Mohri Courant Institute and Google mohri@cims.nyu.edu Motivation Applications: distinct training and test distributions.

More information

Learning Kernels -Tutorial Part III: Theoretical Guarantees.

Learning Kernels -Tutorial Part III: Theoretical Guarantees. Learning Kernels -Tutorial Part III: Theoretical Guarantees. Corinna Cortes Google Research corinna@google.com Mehryar Mohri Courant Institute & Google Research mohri@cims.nyu.edu Afshin Rostami UC Berkeley

More information

Learning Bounds for Importance Weighting

Learning Bounds for Importance Weighting Learning Bounds for Iportance Weighting Corinna Cortes Google Research New York, NY 00 corinna@googleco Yishay Mansour Tel-Aviv University Tel-Aviv 69978, Israel ansour@tauacil Mehryar Mohri Courant Institute

More information

Foundations of Machine Learning

Foundations of Machine Learning Introduction to ML Mehryar Mohri Courant Institute and Google Research mohri@cims.nyu.edu page 1 Logistics Prerequisites: basics in linear algebra, probability, and analysis of algorithms. Workload: about

More information

Advanced Machine Learning

Advanced Machine Learning Advanced Machine Learning Bandit Problems MEHRYAR MOHRI MOHRI@ COURANT INSTITUTE & GOOGLE RESEARCH. Multi-Armed Bandit Problem Problem: which arm of a K-slot machine should a gambler pull to maximize his

More information

Deep Boosting. Joint work with Corinna Cortes (Google Research) Umar Syed (Google Research) COURANT INSTITUTE & GOOGLE RESEARCH.

Deep Boosting. Joint work with Corinna Cortes (Google Research) Umar Syed (Google Research) COURANT INSTITUTE & GOOGLE RESEARCH. Deep Boosting Joint work with Corinna Cortes (Google Research) Umar Syed (Google Research) MEHRYAR MOHRI MOHRI@ COURANT INSTITUTE & GOOGLE RESEARCH. Ensemble Methods in ML Combining several base classifiers

More information

Advanced Machine Learning

Advanced Machine Learning Advanced Machine Learning Deep Boosting MEHRYAR MOHRI MOHRI@ COURANT INSTITUTE & GOOGLE RESEARCH. Outline Model selection. Deep boosting. theory. algorithm. experiments. page 2 Model Selection Problem:

More information

ADANET: adaptive learning of neural networks

ADANET: adaptive learning of neural networks ADANET: adaptive learning of neural networks Joint work with Corinna Cortes (Google Research) Javier Gonzalo (Google Research) Vitaly Kuznetsov (Google Research) Scott Yang (Courant Institute) MEHRYAR

More information

Authors: John Blitzer, Koby Crammer, Alex Kulesza, Fernando Pereira and Jennifer Wortman (University of Pennsylvania)

Authors: John Blitzer, Koby Crammer, Alex Kulesza, Fernando Pereira and Jennifer Wortman (University of Pennsylvania) Learning Bouds for Domain Adaptation Authors: John Blitzer, Koby Crammer, Alex Kulesza, Fernando Pereira and Jennifer Wortman (University of Pennsylvania) Presentation by: Afshin Rostamizadeh (New York

More information

Advanced Machine Learning

Advanced Machine Learning Advanced Machine Learning Domain Adaptation MEHRYAR MOHRI MOHRI@ COURANT INSTITUTE & GOOGLE RESEARCH. Non-Ideal World time real world ideal domain sampling page 2 Outline Domain adaptation. Multiple-source

More information

Learning with Rejection

Learning with Rejection Learning with Rejection Corinna Cortes 1, Giulia DeSalvo 2, and Mehryar Mohri 2,1 1 Google Research, 111 8th Avenue, New York, NY 2 Courant Institute of Mathematical Sciences, 251 Mercer Street, New York,

More information

Does Unlabeled Data Help?

Does Unlabeled Data Help? Does Unlabeled Data Help? Worst-case Analysis of the Sample Complexity of Semi-supervised Learning. Ben-David, Lu and Pal; COLT, 2008. Presentation by Ashish Rastogi Courant Machine Learning Seminar. Outline

More information

arxiv: v1 [cs.lg] 19 May 2008

arxiv: v1 [cs.lg] 19 May 2008 Sample Selection Bias Correction Theory Corinna Cortes 1, Mehryar Mohri 2,1, Michael Riley 1, and Afshin Rostamizadeh 2 arxiv:0805.2775v1 [cs.lg] 19 May 2008 1 Google Research, 76 Ninth Avenue, New York,

More information

Introduction to Machine Learning Lecture 11. Mehryar Mohri Courant Institute and Google Research

Introduction to Machine Learning Lecture 11. Mehryar Mohri Courant Institute and Google Research Introduction to Machine Learning Lecture 11 Mehryar Mohri Courant Institute and Google Research mohri@cims.nyu.edu Boosting Mehryar Mohri - Introduction to Machine Learning page 2 Boosting Ideas Main idea:

More information

Generalization, Overfitting, and Model Selection

Generalization, Overfitting, and Model Selection Generalization, Overfitting, and Model Selection Sample Complexity Results for Supervised Classification Maria-Florina (Nina) Balcan 10/03/2016 Two Core Aspects of Machine Learning Algorithm Design. How

More information

Introduction to Machine Learning Lecture 14. Mehryar Mohri Courant Institute and Google Research

Introduction to Machine Learning Lecture 14. Mehryar Mohri Courant Institute and Google Research Introduction to Machine Learning Lecture 14 Mehryar Mohri Courant Institute and Google Research mohri@cims.nyu.edu Density Estimation Maxent Models 2 Entropy Definition: the entropy of a random variable

More information

Weighted Finite-State Transducers in Computational Biology

Weighted Finite-State Transducers in Computational Biology Weighted Finite-State Transducers in Computational Biology Mehryar Mohri Courant Institute of Mathematical Sciences mohri@cims.nyu.edu Joint work with Corinna Cortes (Google Research). 1 This Tutorial

More information

ECE521 Lectures 9 Fully Connected Neural Networks

ECE521 Lectures 9 Fully Connected Neural Networks ECE521 Lectures 9 Fully Connected Neural Networks Outline Multi-class classification Learning multi-layer neural networks 2 Measuring distance in probability space We learnt that the squared L2 distance

More information

Machine Learning. VC Dimension and Model Complexity. Eric Xing , Fall 2015

Machine Learning. VC Dimension and Model Complexity. Eric Xing , Fall 2015 Machine Learning 10-701, Fall 2015 VC Dimension and Model Complexity Eric Xing Lecture 16, November 3, 2015 Reading: Chap. 7 T.M book, and outline material Eric Xing @ CMU, 2006-2015 1 Last time: PAC and

More information

Rademacher Complexity Bounds for Non-I.I.D. Processes

Rademacher Complexity Bounds for Non-I.I.D. Processes Rademacher Complexity Bounds for Non-I.I.D. Processes Mehryar Mohri Courant Institute of Mathematical ciences and Google Research 5 Mercer treet New York, NY 00 mohri@cims.nyu.edu Afshin Rostamizadeh Department

More information

Machine Learning Basics: Maximum Likelihood Estimation

Machine Learning Basics: Maximum Likelihood Estimation Machine Learning Basics: Maximum Likelihood Estimation Sargur N. srihari@cedar.buffalo.edu This is part of lecture slides on Deep Learning: http://www.cedar.buffalo.edu/~srihari/cse676 1 Topics 1. Learning

More information

Foundations of Machine Learning

Foundations of Machine Learning Maximum Entropy Models, Logistic Regression Mehryar Mohri Courant Institute and Google Research mohri@cims.nyu.edu page 1 Motivation Probabilistic models: density estimation. classification. page 2 This

More information

Foundations of Machine Learning On-Line Learning. Mehryar Mohri Courant Institute and Google Research

Foundations of Machine Learning On-Line Learning. Mehryar Mohri Courant Institute and Google Research Foundations of Machine Learning On-Line Learning Mehryar Mohri Courant Institute and Google Research mohri@cims.nyu.edu Motivation PAC learning: distribution fixed over time (training and test). IID assumption.

More information

Structured Prediction

Structured Prediction Structured Prediction Ningshan Zhang Advanced Machine Learning, Spring 2016 Outline Ensemble Methods for Structured Prediction[1] On-line learning Boosting AGeneralizedKernelApproachtoStructuredOutputLearning[2]

More information

Lecture Support Vector Machine (SVM) Classifiers

Lecture Support Vector Machine (SVM) Classifiers Introduction to Machine Learning Lecturer: Amir Globerson Lecture 6 Fall Semester Scribe: Yishay Mansour 6.1 Support Vector Machine (SVM) Classifiers Classification is one of the most important tasks in

More information

Online Learning for Time Series Prediction

Online Learning for Time Series Prediction Online Learning for Time Series Prediction Joint work with Vitaly Kuznetsov (Google Research) MEHRYAR MOHRI MOHRI@ COURANT INSTITUTE & GOOGLE RESEARCH. Motivation Time series prediction: stock values.

More information

Structured Prediction Theory and Algorithms

Structured Prediction Theory and Algorithms Structured Prediction Theory and Algorithms Joint work with Corinna Cortes (Google Research) Vitaly Kuznetsov (Google Research) Scott Yang (Courant Institute) MEHRYAR MOHRI MOHRI@ COURANT INSTITUTE & GOOGLE

More information

Impossibility Theorems for Domain Adaptation

Impossibility Theorems for Domain Adaptation Shai Ben-David and Teresa Luu Tyler Lu Dávid Pál School of Computer Science University of Waterloo Waterloo, ON, CAN {shai,t2luu}@cs.uwaterloo.ca Dept. of Computer Science University of Toronto Toronto,

More information

The PAC Learning Framework -II

The PAC Learning Framework -II The PAC Learning Framework -II Prof. Dan A. Simovici UMB 1 / 1 Outline 1 Finite Hypothesis Space - The Inconsistent Case 2 Deterministic versus stochastic scenario 3 Bayes Error and Noise 2 / 1 Outline

More information

i=1 Pr(X 1 = i X 2 = i). Notice each toss is independent and identical, so we can write = 1/6

i=1 Pr(X 1 = i X 2 = i). Notice each toss is independent and identical, so we can write = 1/6 Mehryar Mohri Foundations of Machine Learning Courant Institute of Mathematical Sciences Homework assignment 1 - solution Credit: Ashish Rastogi, Afshin Rostamizadeh Ameet Talwalkar, and Eugene Weinstein.

More information

Generalization and Overfitting

Generalization and Overfitting Generalization and Overfitting Model Selection Maria-Florina (Nina) Balcan February 24th, 2016 PAC/SLT models for Supervised Learning Data Source Distribution D on X Learning Algorithm Expert / Oracle

More information

Boosting Ensembles of Structured Prediction Rules

Boosting Ensembles of Structured Prediction Rules Boosting Ensembles of Structured Prediction Rules Corinna Cortes Google Research 76 Ninth Avenue New York, NY 10011 corinna@google.com Vitaly Kuznetsov Courant Institute 251 Mercer Street New York, NY

More information

i=1 cosn (x 2 i y2 i ) over RN R N. cos y sin x

i=1 cosn (x 2 i y2 i ) over RN R N. cos y sin x Mehryar Mohri Foundations of Machine Learning Courant Institute of Mathematical Sciences Homework assignment 3 November 16, 017 Due: Dec 01, 017 A. Kernels Show that the following kernels K are PDS: 1.

More information

Estimating Symmetric Distribution Properties Maximum Likelihood Strikes Back!

Estimating Symmetric Distribution Properties Maximum Likelihood Strikes Back! Estimating Symmetric Distribution Properties Maximum Likelihood Strikes Back! Jayadev Acharya Cornell Hirakendu Das Yahoo Alon Orlitsky UCSD Ananda Suresh Google Shannon Channel March 5, 2018 Based on:

More information

Time Series Prediction & Online Learning

Time Series Prediction & Online Learning Time Series Prediction & Online Learning Joint work with Vitaly Kuznetsov (Google Research) MEHRYAR MOHRI MOHRI@ COURANT INSTITUTE & GOOGLE RESEARCH. Motivation Time series prediction: stock values. earthquakes.

More information

PATTERN RECOGNITION AND MACHINE LEARNING

PATTERN RECOGNITION AND MACHINE LEARNING PATTERN RECOGNITION AND MACHINE LEARNING Chapter 1. Introduction Shuai Huang April 21, 2014 Outline 1 What is Machine Learning? 2 Curve Fitting 3 Probability Theory 4 Model Selection 5 The curse of dimensionality

More information

Computational Learning Theory: Probably Approximately Correct (PAC) Learning. Machine Learning. Spring The slides are mainly from Vivek Srikumar

Computational Learning Theory: Probably Approximately Correct (PAC) Learning. Machine Learning. Spring The slides are mainly from Vivek Srikumar Computational Learning Theory: Probably Approximately Correct (PAC) Learning Machine Learning Spring 2018 The slides are mainly from Vivek Srikumar 1 This lecture: Computational Learning Theory The Theory

More information

Structural Maxent Models

Structural Maxent Models Corinna Cortes Google Research, 111 8th Avenue, New York, NY 10011 Vitaly Kuznetsov Courant Institute of Mathematical Sciences, 251 Mercer Street, New York, NY 10012 Mehryar Mohri Courant Institute and

More information

EE5319R: Problem Set 3 Assigned: 24/08/16, Due: 31/08/16

EE5319R: Problem Set 3 Assigned: 24/08/16, Due: 31/08/16 EE539R: Problem Set 3 Assigned: 24/08/6, Due: 3/08/6. Cover and Thomas: Problem 2.30 (Maimum Entropy): Solution: We are required to maimize H(P X ) over all distributions P X on the non-negative integers

More information

Computational and Statistical Learning theory

Computational and Statistical Learning theory Computational and Statistical Learning theory Problem set 2 Due: January 31st Email solutions to : karthik at ttic dot edu Notation : Input space : X Label space : Y = {±1} Sample : (x 1, y 1,..., (x n,

More information

Statistical Learning Reading Assignments

Statistical Learning Reading Assignments Statistical Learning Reading Assignments S. Gong et al. Dynamic Vision: From Images to Face Recognition, Imperial College Press, 2001 (Chapt. 3, hard copy). T. Evgeniou, M. Pontil, and T. Poggio, "Statistical

More information

Advanced Machine Learning

Advanced Machine Learning Advanced Machine Learning Learning Kernels MEHRYAR MOHRI MOHRI@ COURANT INSTITUTE & GOOGLE RESEARCH. Outline Kernel methods. Learning kernels scenario. learning bounds. algorithms. page 2 Machine Learning

More information

Domain Adaptation with Multiple Sources

Domain Adaptation with Multiple Sources Domain Adaptation with Multiple Sources Yishay Mansour Google Research and Tel Aviv Univ. mansour@tau.ac.il Mehryar Mohri Courant Institute and Google Research mohri@cims.nyu.edu Afshin Rostamizadeh Courant

More information

The definitions and notation are those introduced in the lectures slides. R Ex D [h

The definitions and notation are those introduced in the lectures slides. R Ex D [h Mehryar Mohri Foundations of Machine Learning Courant Institute of Mathematical Sciences Homework assignment 2 October 04, 2016 Due: October 18, 2016 A. Rademacher complexity The definitions and notation

More information

Machine Learning Lecture 7

Machine Learning Lecture 7 Course Outline Machine Learning Lecture 7 Fundamentals (2 weeks) Bayes Decision Theory Probability Density Estimation Statistical Learning Theory 23.05.2016 Discriminative Approaches (5 weeks) Linear Discriminant

More information

Nearest neighbor classification in metric spaces: universal consistency and rates of convergence

Nearest neighbor classification in metric spaces: universal consistency and rates of convergence Nearest neighbor classification in metric spaces: universal consistency and rates of convergence Sanjoy Dasgupta University of California, San Diego Nearest neighbor The primeval approach to classification.

More information

Algorithms for Learning Kernels Based on Centered Alignment

Algorithms for Learning Kernels Based on Centered Alignment Journal of Machine Learning Research 3 (2012) XX-XX Submitted 01/11; Published 03/12 Algorithms for Learning Kernels Based on Centered Alignment Corinna Cortes Google Research 76 Ninth Avenue New York,

More information

Lecture Learning infinite hypothesis class via VC-dimension and Rademacher complexity;

Lecture Learning infinite hypothesis class via VC-dimension and Rademacher complexity; CSCI699: Topics in Learning and Game Theory Lecture 2 Lecturer: Ilias Diakonikolas Scribes: Li Han Today we will cover the following 2 topics: 1. Learning infinite hypothesis class via VC-dimension and

More information

1 Kernel methods & optimization

1 Kernel methods & optimization Machine Learning Class Notes 9-26-13 Prof. David Sontag 1 Kernel methods & optimization One eample of a kernel that is frequently used in practice and which allows for highly non-linear discriminant functions

More information

Learning Theory. Piyush Rai. CS5350/6350: Machine Learning. September 27, (CS5350/6350) Learning Theory September 27, / 14

Learning Theory. Piyush Rai. CS5350/6350: Machine Learning. September 27, (CS5350/6350) Learning Theory September 27, / 14 Learning Theory Piyush Rai CS5350/6350: Machine Learning September 27, 2011 (CS5350/6350) Learning Theory September 27, 2011 1 / 14 Why Learning Theory? We want to have theoretical guarantees about our

More information

Introduction to Machine Learning Lecture 13. Mehryar Mohri Courant Institute and Google Research

Introduction to Machine Learning Lecture 13. Mehryar Mohri Courant Institute and Google Research Introduction to Machine Learning Lecture 13 Mehryar Mohri Courant Institute and Google Research mohri@cims.nyu.edu Multi-Class Classification Mehryar Mohri - Introduction to Machine Learning page 2 Motivation

More information

Machine Learning

Machine Learning Machine Learning 10-601 Tom M. Mitchell Machine Learning Department Carnegie Mellon University October 11, 2012 Today: Computational Learning Theory Probably Approximately Coorrect (PAC) learning theorem

More information

Lecture Slides for INTRODUCTION TO. Machine Learning. By: Postedited by: R.

Lecture Slides for INTRODUCTION TO. Machine Learning. By:  Postedited by: R. Lecture Slides for INTRODUCTION TO Machine Learning By: alpaydin@boun.edu.tr http://www.cmpe.boun.edu.tr/~ethem/i2ml Postedited by: R. Basili Learning a Class from Examples Class C of a family car Prediction:

More information

Perceptron Mistake Bounds

Perceptron Mistake Bounds Perceptron Mistake Bounds Mehryar Mohri, and Afshin Rostamizadeh Google Research Courant Institute of Mathematical Sciences Abstract. We present a brief survey of existing mistake bounds and introduce

More information

Relative Density-Ratio Estimation for Robust Distribution Comparison

Relative Density-Ratio Estimation for Robust Distribution Comparison Neural Computation, vol.5, no.5, pp.34 37, 3. Relative Density-Ratio Estimation for Robust Distribution Comparison Makoto Yamada NTT Communication Science Laboratories, NTT Corporation, Japan. yamada.makoto@lab.ntt.co.jp

More information

Machine Learning: Chenhao Tan University of Colorado Boulder LECTURE 9

Machine Learning: Chenhao Tan University of Colorado Boulder LECTURE 9 Machine Learning: Chenhao Tan University of Colorado Boulder LECTURE 9 Slides adapted from Jordan Boyd-Graber Machine Learning: Chenhao Tan Boulder 1 of 39 Recap Supervised learning Previously: KNN, naïve

More information

Kernel Methods. Lecture 4: Maximum Mean Discrepancy Thanks to Karsten Borgwardt, Malte Rasch, Bernhard Schölkopf, Jiayuan Huang, Arthur Gretton

Kernel Methods. Lecture 4: Maximum Mean Discrepancy Thanks to Karsten Borgwardt, Malte Rasch, Bernhard Schölkopf, Jiayuan Huang, Arthur Gretton Kernel Methods Lecture 4: Maximum Mean Discrepancy Thanks to Karsten Borgwardt, Malte Rasch, Bernhard Schölkopf, Jiayuan Huang, Arthur Gretton Alexander J. Smola Statistical Machine Learning Program Canberra,

More information

Speech Recognition Lecture 8: Expectation-Maximization Algorithm, Hidden Markov Models.

Speech Recognition Lecture 8: Expectation-Maximization Algorithm, Hidden Markov Models. Speech Recognition Lecture 8: Expectation-Maximization Algorithm, Hidden Markov Models. Mehryar Mohri Courant Institute and Google Research mohri@cims.nyu.com This Lecture Expectation-Maximization (EM)

More information

The Perceptron Algorithm

The Perceptron Algorithm The Perceptron Algorithm Machine Learning Spring 2018 The slides are mainly from Vivek Srikumar 1 Outline The Perceptron Algorithm Perceptron Mistake Bound Variants of Perceptron 2 Where are we? The Perceptron

More information

Machine Learning for NLP

Machine Learning for NLP Machine Learning for NLP Uppsala University Department of Linguistics and Philology Slides borrowed from Ryan McDonald, Google Research Machine Learning for NLP 1(50) Introduction Linear Classifiers Classifiers

More information

Lecture 7 Introduction to Statistical Decision Theory

Lecture 7 Introduction to Statistical Decision Theory Lecture 7 Introduction to Statistical Decision Theory I-Hsiang Wang Department of Electrical Engineering National Taiwan University ihwang@ntu.edu.tw December 20, 2016 1 / 55 I-Hsiang Wang IT Lecture 7

More information

Convolution and Pooling as an Infinitely Strong Prior

Convolution and Pooling as an Infinitely Strong Prior Convolution and Pooling as an Infinitely Strong Prior Sargur Srihari srihari@buffalo.edu This is part of lecture slides on Deep Learning: http://www.cedar.buffalo.edu/~srihari/cse676 1 Topics in Convolutional

More information

Learning Weighted Automata

Learning Weighted Automata Learning Weighted Automata Joint work with Borja Balle (Amazon Research) MEHRYAR MOHRI MOHRI@ COURANT INSTITUTE & GOOGLE RESEARCH. Weighted Automata (WFAs) page 2 Motivation Weighted automata (WFAs): image

More information

Machine Learning. Lecture 9: Learning Theory. Feng Li.

Machine Learning. Lecture 9: Learning Theory. Feng Li. Machine Learning Lecture 9: Learning Theory Feng Li fli@sdu.edu.cn https://funglee.github.io School of Computer Science and Technology Shandong University Fall 2018 Why Learning Theory How can we tell

More information

Machine Learning

Machine Learning Machine Learning 10-601 Tom M. Mitchell Machine Learning Department Carnegie Mellon University October 11, 2012 Today: Computational Learning Theory Probably Approximately Coorrect (PAC) learning theorem

More information

CS340 Machine learning Lecture 5 Learning theory cont'd. Some slides are borrowed from Stuart Russell and Thorsten Joachims

CS340 Machine learning Lecture 5 Learning theory cont'd. Some slides are borrowed from Stuart Russell and Thorsten Joachims CS340 Machine learning Lecture 5 Learning theory cont'd Some slides are borrowed from Stuart Russell and Thorsten Joachims Inductive learning Simplest form: learn a function from examples f is the target

More information

Almost k-wise independence versus k-wise independence

Almost k-wise independence versus k-wise independence Almost k-wise independence versus k-wise independence Noga Alon Sackler Faculty of Exact Sciences Tel Aviv University Ramat-Aviv, Israel. nogaa@post.tau.ac.il Yishay Mansour School of Computer Science

More information

arxiv: v4 [cs.it] 17 Oct 2015

arxiv: v4 [cs.it] 17 Oct 2015 Upper Bounds on the Relative Entropy and Rényi Divergence as a Function of Total Variation Distance for Finite Alphabets Igal Sason Department of Electrical Engineering Technion Israel Institute of Technology

More information

Download the following data sets from the UC Irvine ML repository:

Download the following data sets from the UC Irvine ML repository: Mehryar Mohri Introduction to Machine Learning Courant Institute of Mathematical Sciences Homework assignment 2 Solution (written by Andres Muñoz) Perceptron Algorithm Download the following data sets

More information

MUTUAL INFORMATION (MI) specifies the level of

MUTUAL INFORMATION (MI) specifies the level of IEEE TRANSACTIONS ON SIGNAL PROCESSING, VOL. 58, NO. 7, JULY 2010 3497 Nonproduct Data-Dependent Partitions for Mutual Information Estimation: Strong Consistency and Applications Jorge Silva, Member, IEEE,

More information

Linear classifiers selecting hyperplane maximizing separation margin between classes (large margin classifiers)

Linear classifiers selecting hyperplane maximizing separation margin between classes (large margin classifiers) Support vector machines In a nutshell Linear classifiers selecting hyperplane maximizing separation margin between classes (large margin classifiers) Solution only depends on a small subset of training

More information

IFT Lecture 7 Elements of statistical learning theory

IFT Lecture 7 Elements of statistical learning theory IFT 6085 - Lecture 7 Elements of statistical learning theory This version of the notes has not yet been thoroughly checked. Please report any bugs to the scribes or instructor. Scribe(s): Brady Neal and

More information

Application of Information Theory, Lecture 7. Relative Entropy. Handout Mode. Iftach Haitner. Tel Aviv University.

Application of Information Theory, Lecture 7. Relative Entropy. Handout Mode. Iftach Haitner. Tel Aviv University. Application of Information Theory, Lecture 7 Relative Entropy Handout Mode Iftach Haitner Tel Aviv University. December 1, 2015 Iftach Haitner (TAU) Application of Information Theory, Lecture 7 December

More information

Foundations of Machine Learning Multi-Class Classification. Mehryar Mohri Courant Institute and Google Research

Foundations of Machine Learning Multi-Class Classification. Mehryar Mohri Courant Institute and Google Research Foundations of Machine Learning Multi-Class Classification Mehryar Mohri Courant Institute and Google Research mohri@cims.nyu.edu Motivation Real-world problems often have multiple classes: text, speech,

More information

PAC-learning, VC Dimension and Margin-based Bounds

PAC-learning, VC Dimension and Margin-based Bounds More details: General: http://www.learning-with-kernels.org/ Example of more complex bounds: http://www.research.ibm.com/people/t/tzhang/papers/jmlr02_cover.ps.gz PAC-learning, VC Dimension and Margin-based

More information

Computational Learning Theory

Computational Learning Theory Computational Learning Theory Pardis Noorzad Department of Computer Engineering and IT Amirkabir University of Technology Ordibehesht 1390 Introduction For the analysis of data structures and algorithms

More information

November 13, 2018 MAT186 Week 8 Justin Ko

November 13, 2018 MAT186 Week 8 Justin Ko 1 Mean Value Theorem Theorem 1 (Mean Value Theorem). Let f be a continuous on [a, b] and differentiable on (a, b). There eists a c (a, b) such that f f(b) f(a) (c) =. b a Eample 1: The Mean Value Theorem

More information

Maximum Mean Discrepancy

Maximum Mean Discrepancy Maximum Mean Discrepancy Thanks to Karsten Borgwardt, Malte Rasch, Bernhard Schölkopf, Jiayuan Huang, Arthur Gretton Alexander J. Smola Statistical Machine Learning Program Canberra, ACT 0200 Australia

More information

Mehryar Mohri Foundations of Machine Learning Courant Institute of Mathematical Sciences Homework assignment 3 April 5, 2013 Due: April 19, 2013

Mehryar Mohri Foundations of Machine Learning Courant Institute of Mathematical Sciences Homework assignment 3 April 5, 2013 Due: April 19, 2013 Mehryar Mohri Foundations of Machine Learning Courant Institute of Mathematical Sciences Homework assignment 3 April 5, 2013 Due: April 19, 2013 A. Kernels 1. Let X be a finite set. Show that the kernel

More information

Multi-Observation Elicitation

Multi-Observation Elicitation Multi-Observation Elicitation Sebastian Casalaina-Martin Rafael Frongillo Tom Morgan Bo Waggoner Colorado Colorado Harvard UPenn July 2017 1 / 11 Background: Properties of distributions Property or statistic

More information

Lecture 21: Minimax Theory

Lecture 21: Minimax Theory Lecture : Minimax Theory Akshay Krishnamurthy akshay@cs.umass.edu November 8, 07 Recap In the first part of the course, we spent the majority of our time studying risk minimization. We found many ways

More information

Rademacher Bounds for Non-i.i.d. Processes

Rademacher Bounds for Non-i.i.d. Processes Rademacher Bounds for Non-i.i.d. Processes Afshin Rostamizadeh Joint work with: Mehryar Mohri Background Background Generalization Bounds - How well can we estimate an algorithm s true performance based

More information

Generalization bounds

Generalization bounds Advanced Course in Machine Learning pring 200 Generalization bounds Handouts are jointly prepared by hie Mannor and hai halev-hwartz he problem of characterizing learnability is the most basic question

More information

Penalized Squared Error and Likelihood: Risk Bounds and Fast Algorithms

Penalized Squared Error and Likelihood: Risk Bounds and Fast Algorithms university-logo Penalized Squared Error and Likelihood: Risk Bounds and Fast Algorithms Andrew Barron Cong Huang Xi Luo Department of Statistics Yale University 2008 Workshop on Sparsity in High Dimensional

More information

Foundations of Machine Learning Ranking. Mehryar Mohri Courant Institute and Google Research

Foundations of Machine Learning Ranking. Mehryar Mohri Courant Institute and Google Research Foundations of Machine Learning Ranking Mehryar Mohri Courant Institute and Google Research mohri@cims.nyu.edu Motivation Very large data sets: too large to display or process. limited resources, need

More information

Importance Reweighting Using Adversarial-Collaborative Training

Importance Reweighting Using Adversarial-Collaborative Training Importance Reweighting Using Adversarial-Collaborative Training Yifan Wu yw4@andrew.cmu.edu Tianshu Ren tren@andrew.cmu.edu Lidan Mu lmu@andrew.cmu.edu Abstract We consider the problem of reweighting a

More information

Decision Tree Learning Lecture 2

Decision Tree Learning Lecture 2 Machine Learning Coms-4771 Decision Tree Learning Lecture 2 January 28, 2008 Two Types of Supervised Learning Problems (recap) Feature (input) space X, label (output) space Y. Unknown distribution D over

More information

Introduction to Statistical Learning Theory

Introduction to Statistical Learning Theory Introduction to Statistical Learning Theory In the last unit we looked at regularization - adding a w 2 penalty. We add a bias - we prefer classifiers with low norm. How to incorporate more complicated

More information

MATH 220 solution to homework 1

MATH 220 solution to homework 1 MATH solution to homework Problem. Define z(s = u( + s, y + s, then z (s = u u ( + s, y + s + y ( + s, y + s = e y, z( y = u( y, = f( y, u(, y = z( = z( y + y If we prescribe the data u(, = f(, then z

More information

Math 180A. Lecture 16 Friday May 7 th. Expectation. Recall the three main probability density functions so far (1) Uniform (2) Exponential.

Math 180A. Lecture 16 Friday May 7 th. Expectation. Recall the three main probability density functions so far (1) Uniform (2) Exponential. Math 8A Lecture 6 Friday May 7 th Epectation Recall the three main probability density functions so far () Uniform () Eponential (3) Power Law e, ( ), Math 8A Lecture 6 Friday May 7 th Epectation Eample

More information

The sample complexity of agnostic learning with deterministic labels

The sample complexity of agnostic learning with deterministic labels The sample complexity of agnostic learning with deterministic labels Shai Ben-David Cheriton School of Computer Science University of Waterloo Waterloo, ON, N2L 3G CANADA shai@uwaterloo.ca Ruth Urner College

More information

Machine Learning Basics Lecture 4: SVM I. Princeton University COS 495 Instructor: Yingyu Liang

Machine Learning Basics Lecture 4: SVM I. Princeton University COS 495 Instructor: Yingyu Liang Machine Learning Basics Lecture 4: SVM I Princeton University COS 495 Instructor: Yingyu Liang Review: machine learning basics Math formulation Given training data x i, y i : 1 i n i.i.d. from distribution

More information

Asymptotic Active Learning

Asymptotic Active Learning Asymptotic Active Learning Maria-Florina Balcan Eyal Even-Dar Steve Hanneke Michael Kearns Yishay Mansour Jennifer Wortman Abstract We describe and analyze a PAC-asymptotic model for active learning. We

More information

ECS171: Machine Learning

ECS171: Machine Learning ECS171: Machine Learning Lecture 6: Training versus Testing (LFD 2.1) Cho-Jui Hsieh UC Davis Jan 29, 2018 Preamble to the theory Training versus testing Out-of-sample error (generalization error): What

More information

VC dimension and Model Selection

VC dimension and Model Selection VC dimension and Model Selection Overview PAC model: review VC dimension: Definition Examples Sample: Lower bound Upper bound!!! Model Selection Introduction to Machine Learning 2 PAC model: Setting A

More information

Understanding Generalization Error: Bounds and Decompositions

Understanding Generalization Error: Bounds and Decompositions CIS 520: Machine Learning Spring 2018: Lecture 11 Understanding Generalization Error: Bounds and Decompositions Lecturer: Shivani Agarwal Disclaimer: These notes are designed to be a supplement to the

More information

Big Data Analytics. Special Topics for Computer Science CSE CSE Feb 24

Big Data Analytics. Special Topics for Computer Science CSE CSE Feb 24 Big Data Analytics Special Topics for Computer Science CSE 4095-001 CSE 5095-005 Feb 24 Fei Wang Associate Professor Department of Computer Science and Engineering fei_wang@uconn.edu Prediction III Goal

More information