EE698R: Advanced Topics in Machine Learning (Spring 2024)

with focus on Generative AI and Trustworthy AI for audio and physical sciences.

Units: 3-0-0-0-9 (3 hours lecture; total 9 credits)
Class timings: MW 14:00-15:10
Instructor: Vipul Arora
Office hours: After the class on Monday and Wednesday

For Registration

• I am planning to have around 50 UGs and rest all PGs.
• No limit on the number of PGs.
• For UGs:
  • First come first serve.
  • Please do not email me. Apply via Pingala.

Course Objectives:

This course aims at introducing the students to advanced topics in machine learning (ML). The main focus will be on Generative machine learning and Trustworthy AI. The lectures will focus on mathematical principles, and there will be coding based assignments/project for implementation.

Pre-requisites:

• Basic course on machine learning (EE698V, EE603A or equivalent). Lecture videos
• Basics of Programming (ESc101, EE698K or equivalent)

The course will need a strong background in linear algebra and probability theory.

Topics:

• Basics of probability theory
• Basics of discrete time signal processing
• Audio as time series
  • speech recognition
  • audio event detection
  • music transcription
• Time series models
  • RNN, seq2seq, wavnet
  • Forward backward algo, CTC, RNN-transducers
• Trustworthy AI
• Uncertainty Estimation
  • confidence calibration
    • confidence estimation in classification
    • confidence estimation in regression
      • confidence intervals
  • Epistemic and aleatoric uncertainty
  • Evidential learning
• Explainable AI
  • LIME, SHAP, Grad-CAM
• Sampling
  • VAEs, GANs, diffusion models
• Monte Carlo Sampling
  • MCMC methods, HMC, MH algorithm
  • Normalizing flows
  • Score-based models
• Applications in audio and physical sciences

Grading Scheme

• Quizzes/Assignments - 30%
• Mid-semester Exam – 30%
• End-semester - 40%

Minimum attendance of 80% is needed to pass the course.

Plagiarism Penalty:

As heavy as possible. Zero-tolerance policy.

References:

This course will take excerpts from some standard books on machine learning and signal processing. But it will largely be based on articles and research papers in ML and SP conferences (e.g., NeurIPS, ICML, ICLR, Interspeech, ICASSP, etc.) and journals (e.g., IEEE TASLP, JMLR, IEEE PAMI, etc.).

Books:

• “Pattern Recognition and Machine Learning”, C.M. Bishop, 2nd Edition, Springer, 2011.
• “Bayesian reasoning and machine learning”, D. Barber, Cambridge University Press, 2012.
• “Deep Learning”, I. Goodfellow, Y, Bengio, A. Courville, MIT Press, 2016.