CS6140 Machine Learning, Spring 2014

about CS6140            home            schedule       grades      final project     forum

*subject to change
Jan 6
Basic concepts, prerequisites

Matrix Data
Decision rules /heuristics

Decision trees
Regression Trees

Training and testing
DHS 8.1-8.5
Other Decision Tree notes
HW1 due Jan 20

Jan 13
Linear regression, Intro to Linear Algebra Methods
Normal Equations

Cross validation
evaluation: accuracy/error, ROC, F1, precision, recall

DHS ch 5
KMPP ch 7,8

Jan 20
NO CLASS (MLK day), discussion on Piazza
Gradient Descent, Intro to numerical methods
Regression With GD
Logistic regression

DHS ch 5 HW2 due Feb 10
Jan 27
Neural Networks
Other slides (Mitchell)
Mitchell chapter on NN
DHS ch6

Feb 3
Probabilities Primer, Bayesian statistics
Generative Methods
Fitting parameters to data, Maximum Likelihood

DHS ch 2,3
KMPP ch 2,3
Andrew Ng Lecture Notes

Feb 10
Gaussian Discriminant Analysis
Naive Bayes
EM algorithm, Gaussian mixtures

DHS ch 2,3
KMPP ch 3,4
EM code, example
EM wikipedia notes , EM Berkeley notes  ,  EM other notes
EM slides set1 , set2

HW3 due Mar 10
Feb 17
Graphical Models, factor graphs (Bayes Nets)


Graphical models - Bishop slides
KMPP ch 20 , KMPP ch 11(EM)

Feb 24
Boosting, AdaBoost
Online Learning,  Hedge Theory

DHS ch 9
KMPP ch 16
boosting testing error proof
AdaBoost/Hedge paper
boosting preferences/ranking
HW4 due Mar 24
Office hours happens as usual WED, THU, FRI
Additional Office hours MON 3/3 5-7PM

Mar 10
Active Learning
VC dimension, PAC complexity
Multiclass: ECOC

Gradient Boosting
Learning to Rank: RankBoost, LambdaMart

VC dimension slides
DHS ch 9.5
KMPP ch 16.6

KMPP ch 27.6.2
Gradient Boosting (LambdaMart Theory)
LambdaMart Slides
LambdaMart JForestspaper, implementation

Mar 17
Support Vector Machines
Convex Optimization , slides
DHS ch 5.11
KMPP ch 14
Chris Burges SVM tutorial , SVM on wikipedia
SVM and VC dimension

Project Proposal due
For students that choose a project
instead oh HW6

Mar 24
Support Vector Machines
SMO, Kernels (notes scan, some from Andrew Ng) Simplified SMO

Andrew NG SVM notes
kernel slides
HW5 due Apr 7
Mar 31
Locality, similarity
K-Nearest Neighbor

Learning with harmonic functions
Parzen Windows

KMPP ch14, ch1
DHS ch 4

Apr 7
Clustering, KMeans
Clustering Hierarchical

Collaborative Filtering

Mirek Riedewald clustering slides
Clustering: DHS ch 3.8, KMPP ch 12.2
Project Milestone
HW6 due April 22 for students
that choose HW6 instead of a project
Apr 14
Feature Selection
Feature Extraction
Dimmensionality reduction
Principal Component Analysis
(lecture follows DHS)

PCA: DHS ch 10, KMPP ch 25

Apr 22 ?
(exam week)
Project Presentation

Project Due