Reading list
This list is horribly out of date — will be updated soon 🙂
- Amir Beck, /First-Order Methods in Optimization/. A really well-written text bringing together a wealth of material on fundamental optimization theory and first-order convex optimization algorithms.
- Nemirovski's recent update to his classical /Lectures on Modern Convex Optimization/ (PDF). The first three chapters are the natural next step after working through the MOSEK modeling cookbook.
- An Introduction to Optimization on Smooth Manifolds by Nicolas Boumal (current maintainer of PyManOPT).
- Computational Optimal Transport — a well-written introduction to OT for mathematically inclined readers, co-authored by Marco Cuturi.
- Semidefinite approximations of the matrix logarithm — Fawzi, Saunderson, Parrilo. My evolving commentary lives on the blog.
- Distributional Reinforcement Learning with Quantile Regression. For a comprehensive introduction to distributional RL, see here.
- Practical Near Neighbour Search via Group Testing — clever use of Distance-Sensitive Bloom Filters coupled with ideas from group testing. Outperforms FAISS by large factors.
- /Numerical Linear Algebra/ by Trefethen and Bau. The best self-contained introduction to NLA there is — a good way to consolidate before graduating to Matrix Computations as a reference.
- Introduction to Online Convex Optimization — Elad Hazan's text. Online learning is fascinating and I want to dig in properly.
- Convex Optimization: Algorithms and Complexity — a beautiful monograph on the algorithmics of convex optimization.
- Non-Convex Optimization for Machine Learning — Prateek Jain's monograph on broad ideas in non-convex optimization.
- Tengyu Ma's StatML notes — great companion to Shai Shalev-Shwartz's learning-theory text, with material on NTK ideas.
- /A User's Guide to Measure-Theoretic Probability Theory/ by David Pollard. An amazing self-contained tour of MTPT — probably sufficient for a wannabe applied mathematician.