Teaching
Spring 2024
Machine Learning for Signal Processing (ENGR-E 511, CSCI-B 590)
This course focuses on the mathematical foundations of machine learning with applications-oriented assignments in signal processing.
Topics discussed:
- Machine learning topics
- Probability Theory (Binomial, Multinomail, Normal, Poisson distributions, MLE, Bayes)
- Linear Algebra for ML (matrix decomposition, vector calculus)
- Numerical Optimization (gradient descent, Newton’s method, automatic differentiation)
- Dimensionality reduction
- Clustering
- Bayesian classification
- Undirected graphical models (Markov random fields, variational inference, Mean-field approximation)
- Nonlinear methods (spectral clustering, kernel methods)
- Neural networks
- Hidden Markov models
- Support vector machines
- Probabilistic topic modelling
- Adaptive methods
- Signal processing topics
- Fourier transformation
- Audio denoising
- Souce seperation
- Stereo matching
- Image segmentation
- Keyword detection
- Brain signal processing
Prior Teaching Experiences (Associate Instructor or Co-lecturing)
-
Fall 2022, Spring 2023 : Deep Learning Systems - ENGR 533 (AI)
-
Spring 2022, Spring 2021 : Cyber-Physical Systems - ENGR 210 (AI)
-
Fall 2021, Advanced Cyber-Physical Systems - ENGR-321 (AI + Co-conducted lectures)
-
Spring 2019, Internet of Things - ENGR-523 (AI)
-
Fall 2018, Machine Learning for Signal Processing - ENGR 511 (AI)