Machine Learning Fall 2021

CLASS HOURS
Tues: 5:30 – 7:00pm, Thurs: 7:15 – 8:45pm
Location: Saeed-ul-Hassan Auditorium


CONTACT

Instructor: Dr. Arif Mahmood
Contact: arif.mahmood@itu.edu.pk

Teaching Assistant 1: Maria Marrium
Contact: msds20011@itu.edu.pk

Teaching Assistant 2: Sohail Danish
Contact: msds20004@itu.edu.pk

COURSE BASICS
Core Course
Credit Hours: 3
Being offered to PhD, MSDS, MSCS, and BS students
Practical and hands-on approach
4 to 5 programming assignments


PREREQUISITE
Artificial Intelligence, Data Structures,
Probability & Statistics, Linear Algebra and basic Calculus
Programming skills and desire to read & implement.

GRADING POLICY

● 35% Final Exam
● 20% Midterm Exam
● 20% Assignments
● 10% Final Project
● 10% Quizzes
● 05% Homeworks

HONOR CODE

All cases of academic misconduct will be forwarded to the disciplinary committee. All assignments are individual unless explicitly specified by the instructor.

COURSE OUTLINE

Lecture

    Topics covered

      Quiz/ Homework/ Programming Assignments

        1

        Intro to ML and its application

        2

        K- Nearest Neighbor, Gradient Descent Linear Regression

        3

        Lecture: Gradient Descent, Classification Vs
        Regression Tutorial: Plotting functions of one, two
        variables and how to compute gradient descent of one variable signal, Numpy demo

        4

        Lecture: Solving Linear regression using

        Gradient Descent, Logistic regression

        Tutorial: Linear Regression

        Class Quiz 01

        5

        Prior Probabilities, Class-Conditional Probabilities, Posterior Probabilities, Bayes Rule

        Programming Assignment 1 (Linear Regression)

        6

        Gaussian Residuals Models, L2 and L1 regularization, Early stopping

        7

        Non-Linear Decision Surfaces, Intro to Neural Network, Activation functions, forward pass, backward pass

        Class Quiz 02

        8

        Neural Network Backpropagation

        Take Home Quiz 1 + Project
        Idea Submission

        9

        Lecture: Activation Functions, Multi-layer neural network, multi-class classification, Cost function and its gradient descent, Stochastic Gradient descent, Optimizers, regularization, learning curves Tutorial: Basics of Pytorch, Neural Network for Digit classification

        10

        Lecture: Overfitting, Model capacity, How neural networks see? Deep neural networks.
        Tutorial: Deep Neural network for digit
        classification

        Homework 3

        11

        LeNET, AlexNet, Feature detection, sobel filter, Blob detection, Convolution layer, FC Layer
        Programming Assignment 2 (Neural Networks)

        12

        Convolution Class Practice

        13

        Lecture: Training CNN

        Tutorial: Digit classification using CNN

        14

        Understanding ConvNets, What does a neuron do in Convnet? Typical backpropagation, Guided backpropagation, Gradient Ascent

        Class Quiz 03

        15

        Discriminative vs Generative Models, Naive Bayes, Logistic regression,

        16

        Multivariate Gaussian, Gaussian Bayes Classifier, K-means

        Programming Assignment 3
        (Convolution Neural Network)

        17

        Dimensionality Reduction, PCA, Autoencoders

        18

        PCA revision, Graphs basics

        Programming Assignment 4 (PCA)

        19

        Graph Shift Operator

        Class Quiz 04

        Mid Terms

        20

        –

        Homework 4

        21

        Graph basics and Graph Shift operator revision, Graph Signal, Graph Convolution

        Winter Vacations

        22

        Graphs revision, Graph frequency response, Graph frequency response of graph filters

        23

        Decision tree

        24

        Decision tree continued, GANS

        Take Home Quiz 2

        25

        Knowledge Distillation

        26

        Anomaly Detection using dictionaries and Autoencoders

        27

        Weakly supervised video anomaly detection

        28

        Human Pose Estimation
        Class Quiz 5

        29

        Reinforcement Learning
        Project Progress report + Project Term paper

        30

        Image Generators and Pose transfer

        Final Exams