MATH 574M - Statistical Machine Learning and Data Mining

  • First class on 08/22.
  • Classroom changed to HARV 102, starting on 08/29.
  • Course re-opened for registration on 08/29.

    Course Information
    Lectures: Tue. and Thu. 9:30-10:45pm, HARV 102| Syllabus
    Office Hours: Tuesday, Thursday 11-12pm, ENR2 S323. Or by appointment.
    Textbooks: The Element of Statistical Learning:data miming, inference, and prediction Hastie, Tibshirani, and Friedman (2001).
    Reference Books:
  • Principle and Theory for Data Mining and Machine Learning by Clark, Forkoue, Zhang (2009)
  • Pattern Recognition and Neural Networks by B. Ripley (1996)
  • Learning with Kernels by Scholkopf and Smola (2000)
  • The Nature of Statistical Learning Theory by Vapnik (1998)
  • An overview of statistical learning theory, Vapnik (1999)

    Useful Links:
  • Kernel Machines
  • Hastie's Software and Data

    R Resouces:
  • Download R (CRAN)
  • Introduction to R | R for Beginners | R reference card

    Statistics Prerequisites:
  • Basic Topics | Joe Watkins' 363 Notes | Joe Watkins' MATH 464 Notes

    Course Activities
    Week 1 (Aug 21-27) Get Familiar with Software: Intrudction to R R Brief Intro, R Guide For Reginners
    Read Chapter 1: Overview of Data Mining Lecture 1: Introduction
    Supplementary Reading: Data mining and statistics: what is the connection? Friedman (1997) Homework 1. Assigned on Aug 22, due on Sep 4.
    Week 2 (Aug 28-Sep 3) Read Chapter 2: Theory of Supervised Learning Lecture 2: Statistical Decision Theory (I)
    Lecture 3: Statistical Decision Theory (II)
    Week 3 (Sep 4 - Sep 10) Read Chapter 4.2-4.4: Linear Classificaton Methods for Binary Problems Lecture 4: Binary Classification (I): Basics
    Homework 2 Assignment. Assigned on Sep 5, due on Sep 19.
    Homework 2 Solution, Code
    Week 4 (Sep 11 - Sep 17) Supplementary Reading: Choosing Between Logistic Regression and Discriminant Analysis, Press, S. and Wilson, S. (1978) Lecture 5: Binary Classification (II): Logistic Regression and Discriminant Analysis
    Curse of Dimensionality; Linear Binary Classification for High Dimensional Problems Lecture 6: Binary Classification (III): Extension to High Dimensional Classification Problems
    Week 4 (Sep 18 - Sep 24) Read Chapter 4.1: Nonlinear Classification Methods Lecture 7: K nearest neighbor (Knn) methods
    Topic: Introduction to Multiclass Classifiction Lecture 8: Multiclass Classification
    Homework 3 Assignment.Assigned on Sep 19, due on Oct 3. Solution, Code
    Week 5 (Sep 25 - Oct 1) Topic: Nonlinear Discriminant Analysis Lecture 9: QDA and RDA
    Supplementary Reading: LDA for improved large vocabulary continuous speech recognition Lecture 10: PCA
    Week 6 (Oct 2 - Oct 8) Topic: Linear Regression Models Lecture 11: Linear Regression
    Read Chapter 3 : Linear Regression, Supplementary Reading: Linear Model Theory Homework 4 Assignment, Solution, Code. Assigned on Oct 3, due on Oct 24 Solution, Code
    Week 7 (Oct 9 - Oct 15) Read Chapter 3 : Variable Selection for Linear Regression Lecture 12: Variable Selection (I)
    Reading:Regression Shrinkage and Selection via the LASSO,
    Week 8 (Oct 16 - Oct 22) Final Project: Project assigned on Oct 19th, due on Dec 12 Lecture 13: Shrinkage Methdods by LASSO
    Final Project: Suggested Reading List
    Week 9 (Oct 23 - Oct 29) Supplementary Reading: Regularization and variable selection via the elastic net Homework 5, Prostate data set, data info. Assigned on Oct 24th, due on Nov 14.Solution, Code
    Lecture 14: Beyond LASSO, Lecture 14+: Functional ANOVA
    Week 10 (Oct 30 - Nov 5) Read Chapter 8 : Model Tuning and Evaluation Lecture 15: Model Selection and Assessment
    Supplenmentary Reading: Leave-out-one Cross Validation
    Week 11 (Nov 6 - Nov 12) Read Chapter 4 (4.5) Lecture 16: Modern Classification vis Separting Hyperplanes
    Read Chapter 12 Lecture 17: Support Vector Machines
    Week 12 (Nov 13 - 19) Supplementary Reading: The Entire Regularization Path for the Support Vector Machine Lecture 18: Multiclass Support Vector Machines
    Homework 6, assigned on Nov 16, due on Dec 5 .

  • Auditors are expected to attend class regularly and submit homework on the same schedule as the other students.

    Policy on Academic Integrity
  • The University policy on academic integrity is spelled out in UA Code of Student Conduct.

    Students with Disabilities
  • Reasonable accommodations will be made for students with verifiable disabilities.