PH6232 Machine Learning Course

Course Materials

The rough course outline will be as follows,
1. Week1: Basic math of a neural network, Basic python,numpy,pandas
2. Week2: Simple regression neural network, testing it and benchmarking performance
3. Week3: Simple classification, train/test splits and ROC curves
4. Week4: Advanced classification task, Ranking of input variables
5. Week 5: Convolutional neural networks
6. Week 6: Examples of unsupervised learning algorithms (clustering)
7. Week 7: Autoencoders, and generative adversarial networks
8. Week 8: continue previous discussion and assessment
9. Week 9: continue previous discussion and assessment
10.(For Week8 and Week9, we may have a guest lecture)

Preliminary Setup

There are two ways to work,
(a) On your laptop: you need a python+relevant packages setup. The easiest way to do this (on windows or linux) is to have a miniconda setup and then install the requisite packages (python3, numpy, pandas, matplotlib, scikit-learn, tensorflow, keras). But you can have them installed any other way too (pip)
Here you can use a local jupyter notebook, or use simple text files and the terminal. (We have understanding of both these methods, so we can help)
(b) In the cloud: use google collab (which is basically a jupyter notebook in the cloud) We will spend some of the first week on getting the setup to do things, so dont worry if you cant get it setup yourself.

Instructors

1. Prof Sourabh Dube (Instructor)
Email: sdube@iiserpune.ac.in
Office: Main Building, Room # A362

2. Arnab Laha (TA)
Email: laha.arnab@students.iiserpune.ac.in
Office: Main Building, Room # A365

(Ehep website: IISER EHEP)