EE 5250: Machine Learning: A Signal Processing Perspective
(Dual-listed with EE 4250).
Credits: 3. Contact Hours: Discussion 1, Lecture 2.
Prereq: Graduate, Professional or Concurrent Student
Background material review (probability, calculus, linear algebra), Key machine learning tools and techniques. Supervised Learning: Linear Regression, Logistic Regression, Generative algorithms for classification (Gaussian & discrete-valued case; Naive Bayes assumption), Support Vector Machines, Decision trees; Unsupervised Learning: principal components analysis (PCA), robust PCA, clustering; Introduction to Deep Learning and Neural Networks; Basic Learning Theory and Bias-Variance Tradeoff; introduction to key Bayesian estimation concepts (MMSE estimation, Kalman filter, hidden Markov models).