Instructor: David B. Cooper
Office: Barus & Holley Room 318
Phone: 863-2601
E-mail: cooper@lems.brown.edu
Spring Semester, 2008
Graduate Level
Meeting Room: Barus & Holley 194
Meeting Time: Tues., Thurs., 9:00-10:20 AM
Final Examination
Monday, May 12, 2008, 1:00-4:00
Room 194
Open Book.: You may bring your course text book (Duda, Hart, and Stork), and notes that were distributed, and your own notes.
Assignments:
Assignment 1: (Due Tuesday Feb 19)
Chapter 2 in your text
Computer Exercises: 1, 2 on pages 79, 80.
Problems: 1, 2, 10, 18
Assignment 2: (Due Tuesday March 4)
Chapter 2 in your text
Problems: 6; 13; 14; 23; 26; 32; 33; 35
Note: Homework assignments must be completed and handed in on time. Make a photocopy for yourselves
This course is a solid reasonably broad treatment of pattern
recognition, with some application to computer vision. The pattern
recognition topics in this course are fundamental to speech analysis,
computer vision, machine learning, statistical signal processing, human
perception and cognition, data mining, medical diagnosis by computer,
etc. These are problems involving making reliable decisions based on
large complicated data sets in the presence of considerable uncertainty
and noise. Roughly 15% of the course will be devoted to application of
the theory and techniques to estimating free-form 3D shapes and
recognizing them based on unorganized messy 3D data or from images
taken at multi-views. The emphasis here is on how to put a complex
application problem within the pattern recognition framework.
Topics:
Bayesian Decision Theory: A solid treatment of
classification theory in terms of Bayesian costs, decision functions
and the geometry of decision regions for continuous and discrete random
variables. Classification error probabilities and bounds; missing
features; Bayesian belief networks.
Maximum-Likelihood And Bayesian Parameter Estimation, and Bayesian Recognition Using A Priori Partially Unknown Distributions:
General theory; Sufficient statistics; Large sample behavior for
arbitrary distributions; Principal component analysis and
discriminants; EM algorithm.
Nonparametric Recognition: Parzen windows classifiers; K-Nearest-Neighbor classifiers.
Support Vector Machines.
Multilayer Neural Networks: Introduction to feedforward operation and classification; Backpropagation algorithm; Behavior considerations.
Decision Trees: CART (classification and regression trees).
Algorithm-Independent Machine Learning: Resampling for estimating statistics and classifier accuracy --- Bootstrap; Boosting.
Unsupervised Learning And Clustering: Mixture densities
and identifiability; K-Means clustering; Unsupervised Bayesian
learning; Decision-directed approximation; Hierarchical clustering;
Minimum spanning trees.
Applications to estimation and recognition of 3D geometry from 3D range data or from multi-view images.
Course Requirements
A midterm and a final examination, 6 homework assignments, a few MATLAB assignments.
Prerequisite: A solid one semester undergraduate course in
statistics or probability theory, or the equivalent, and some knowledge
of linear algebra.
Texts:
Pattern Classification (2nd ed.), by R. O. Duda, P. E. Hart and D. G. Stork.
Pattern Recognition and Machine Learning (Information Science and Statistics), by Christopher M. Bishop.
Useful Books on Reserve for Course EN252:
Probability, Random Variables and Stochastic Processes, 4th
edition, by A. Papoulis and S.U. Pillai, McGraw Hill, 2002. ISBN
0-07-366011-6. QA273.P2 2002.
Statistical Signal Processing. Detection, Estimation and Time
Series Analysis, by L.L. Scharf, Addison Wesley, 1990. ISBN
0-201-19038-9. TK5102.5.S3528 1990.
Introduction to Statistical Pattern Recognition, 2nd edition, by K. Fukunaga, Academic Press, 1990. ISBN 0-12-269851-7.
Elements of Statistical Learning. Data Mining, Inference, and
Prediction, by T. Hastie, R. Tibshirani, J. Friedman, Springer, 2001.
ISBN 0-387-95284-5. Q325.75.F75 2001.
