Schedule

This schedule is meant as an outline. Depending on progress, material may be added or removed. Also, there will often be interesting tangents to follow.

Date Topics Covered Discussion
1 Jan. 07 – 11     Course Overview & kNN

Cross-validation
Leave-one-out		 dis. 1
2 Jan. 14 – 18     Decision Tree & Naive Bayes

Entropy and Gini impurity

Reduced-Error Pruning
Naive Bayes assumption

PA1 due Jan. 25		 dis. 2
3 Jan. 21 – 25 Linear Regression

Residual Sum of Squares
Nonlinear basis
Regularization

TA1 due Feb. 01		 dis. 3
4 Jan. 28 – Feb. 01 Perceptron & Logistic Regression

Gradient Descent
Surrogate Losses

Multiclass Classification
PA2 due Feb. 08		 dis. 4
5 Feb. 04 – 08 Neural Networks

Backpropagation

Preventing overfitting
TA2 due Feb. 15		 dis. 5
6 Feb. 11 – 15 Convolutional Neural Networks
PA3 due Mar. 10		 dis. 6
7 Feb. 18 – 22 Kernels & Clustering

Mercer Theorem
Kernelizing ML algorithms

K-means clustering		 dis. 7
8 Feb. 25 – Mar. 01 Review for exam
Exam – I 		no discussions
9 Mar. 04 – 08 Support Vector Machines

Linear Programming
Lagrangian Duality

KKT conditions
Dual SVM
TA3 due Mar. 24		 dis. 8
10 Mar. 18 – 22 Boosting & Gaussian Mixture Models
AdaBoost 		dis. 9
11 Mar. 25 – 29 Gaussian Mixture Models

EM algorithm
Density estimation
TA4 due Apr. 19		 dis. 10
12 Apr. 01 – 05 Hidden Markov Models

Markov chains
Viterbi algorithm

PA4 due Apr. 21		 dis. 11
13 Apr. 08 – 12 HMM & PCA

Baum-Welch algorithm
PCA algorithm		 dis. 12
14 Apr. 15 – 19 Reinforcement Learning

E.Brunskill notes

Multi-Armed Bandits
Markov Decision Processes

Bellman's optimality principle		 no discussions
15 Apr. 22 – 26 Review for exam
Exam – II 		no discussions