High-Dimensional Probability and Inference (CS-0209)
Spring 2027 · CS 0209 · Tufts University · 3 credits · Undergraduate
Introduction to the ubiquitous concentration of measure phenomena for sums of many (hence high-dimensions) independent random variables. Scalar and vector-valued Sub-Gaussian and Sub-Exponential distributions, their concentration properties expressed as Hoeffding and Bernstein Inequalities, concentration of operator norms of random matrices, and suprema of stochastic processes with regularity properties via generic chaining. Applications include analysis of spectral clustering for community detection, randomized dimensionality reduction (Johnston-Lindestrauss Lemma) with provable guarantees, and error rates on statistical estimation of covariance matrices for PCA. Recommendations: Undergraduate Linear Algebra and Probability classes
Course codes: CS-0209, CS 0209, CS0209, CS-209, CS 209, CS209, EE-0209, EE 0209, EE0209, EE-209, EE 209, EE209, MATH-0264, MATH 0264, MATH0264, MATH-264, MATH 264, MATH264
- 01-LEC (Lecture) — Shuchin Aeron