STATDIY bookmarks

Convergence proofs for gradient descent, SGD, mini-batching, momentum, subgradient, and proximal methods.

Accessible introduction to quasi-Newton methods and the BFGS inverse-Hessian update.

Chinese explainer deriving NAG from momentum, comparing update formulas, and interpreting acceleration through gradient-change information.

CMU optimization course with slides, scribed notes, and homework on convexity, duality, first-order, proximal, Newton, ADMM, Frank-Wolfe, and stochastic methods.

Large-scale convex optimization course covering gradient, subgradient, proximal, accelerated, splitting, ADMM, conic, and interior-point methods.

Proof-based optimization theory course covering convex optimization, stochastic and variance-reduced methods, nonconvex optimization, saddle-point escape, and minimax optimization.

Optimization-for-ML course covering first-order methods, regret and generalization, mirror descent, variance reduction, Frank-Wolfe, Nesterov acceleration, and second-order methods.

Graduate large-scale optimization course for data science covering smooth and nonsmooth algorithms, proximal methods, ADMM, quasi-Newton, stochastic, variance-reduced, and distributed optimization.

Theory-oriented course on scalable convex optimization for machine learning and high-dimensional statistics, emphasizing problem structure and numerical algorithm complexity.

Chinese video course on convex optimization foundations, convex analysis, duality, and standard algorithmic topics.

Interactive Julia and Pluto course linking programming abstractions, mathematics, data science, stochastic simulation, optimization, climate modeling, and open-source workflows.

Scientific Python course for economics and finance covering Python fundamentals, NumPy, plotting, numerical programming, and computational workflows.

MIT OCW version of the Julia-based computational thinking course with interactive notebooks for modeling, simulation, data analysis, and images.

Scientific programming course for physical scientists covering FORTRAN, C, C++, MATLAB, Mathematica, program design, algorithm development, and verification.

Graduate numerical analysis course focused on accurate and efficient algorithms for numerical linear algebra, floating-point arithmetic, conditioning, stability, and nonlinear optimization.

Repository for newer 18.335 materials and Julia notebooks supporting numerical methods, numerical linear algebra, and related computational exercises.

Open textbook covering probability axioms, random variables, limit theorems, mathematical statistics, random processes, and simulation.

Applied causal inference handbook built around DAGs, potential outcomes, DiD, IV, RDD, matching, and synthetic control.

Broad mathematical statistics text connecting probability models, estimation, testing, regression, ANOVA, and data analysis examples.

Advanced probability reference for measure-theoretic foundations, convergence, conditioning, martingales, and limit theorems.

Rigorous graduate probability text for measure theory, independence, convergence, laws of large numbers, CLT, martingales, and Markov chains.

Probability foundations with an unusually friendly path into sigma-algebras, conditioning, independence, and convergence.

Measure-theoretic probability path through distributions, expectation, independence, convergence, laws of large numbers, and CLT.

Probability course notes on counting, conditional probability, random variables, common distributions, transformations, and the CLT.

Graduate multivariate statistics notes for MANOVA, PCA, factor analysis, canonical correlation, discriminant analysis, and clustering.

Statistical computing toolbox: numerical linear algebra, optimization, simulation, and sampling.

Graduate statistics course on information-theoretic tools for concentration, generalization, minimax lower bounds, online convex optimization, and bandits.

Companion answer reference for checking exercises from the Chinese mathematical statistics text.

Classic statistical learning reference for supervised and unsupervised methods, regularization, trees, boosting, SVMs, and model assessment.

Rigorous learning theory with high-dimensional probability, statistics, and optimisation.

Hands-on data mining course with R, EDA, model selection, regression, PCA, classification, SVMs, trees, and clustering.

Advanced CMU notes on density estimation, regression, sparsity, classifiers, forests, graphical models, causal inference, privacy, and optimal transport.

Matrix-methods bridge from linear algebra to data analysis, signal processing, optimization, and machine learning.

Signal-processing route into statistical learning, spectral methods, filtering, and structured data representations.

Practical transformer NLP book covering tokenizers, Hugging Face workflows, fine-tuning, datasets, and deployment patterns.

Computer vision deep-learning course focused on CNNs, visual recognition, end-to-end training, and practical model-tuning assignments.

LLM foundations and safety course covering interpretability, scaling laws, adversarial robustness, alignment, governance, privacy, watermarking, and evaluation.

Deep RL course with lectures, slides, and assignments from imitation learning through policy gradients, Q-learning, model-based RL, and exploration.

Core RL sequence from MDPs through dynamic programming, model-free methods, value approximation, and policy gradients.

Gym API reference for building and testing RL agent-environment loops, spaces, wrappers, and benchmark environments.

Video lecture entry point for DeepMind-style RL intuition, terminology, and algorithmic framing.

Two-volume analysis reference with rigorous calculus, topology of Euclidean spaces, differentiation, integration, series, and manifolds.

Canonical compact real-analysis text for metric spaces, sequences, continuity, differentiation, Riemann-Stieltjes integration, and function sequences.

Computational mathematics text for error analysis, nonlinear equations, linear systems, interpolation, integration, ODEs, and eigenvalue problems.

Graduate real analysis reference covering measure, integration, signed measures, functional analysis, Hilbert spaces, and Fourier analysis.

Twenty-six real analysis lectures from constructing number systems through compactness, continuity, derivatives, series, and Taylor's theorem.

Comprehensive C programming book for syntax, pointers, memory, arrays, structs, I/O, and systems-level basics.

Practical Linux reference for shell usage, filesystem layout, permissions, networking, process management, and server administration.

Applied Linux administration guide focused on command-line workflows, services, networking, storage, and deployment tasks.

Bottom-up systems text connecting bits, logic gates, machine architecture, assembly, C, memory, and OS-level abstractions.

Practical computing course covering shell, editors, Git, debugging, profiling, automation, security, and other tools worth learning early.

Professional certificate sequence teaching C programming, pointers, memory management, compilation, Linux command-line tools, and systems programming practice.

Focused C/Linux course on compiling, libraries, makefiles, command-line tools, and practical systems programming workflows.

Research computing lesson collection for Unix shell, Git, Python, R, reproducible workflows, automation, and collaborative scientific programming.

Genre-based academic writing guide for non-native English speakers, organized around the sections of scientific papers.

Chinese translation of Donoho's essay framing data science through statistics, computation, and real-world data analysis.

Breiman interview for revisiting the statistical modeling cultures debate and his path into data-driven modeling.

Discussion of data modeling versus algorithmic modeling and what changed after Breiman's critique.

Hastie and Tibshirani conversation revisiting The Two Cultures from statistical learning practice.

Academic workflow checklists for reading, writing, reviewing, giving talks, and navigating Stat-ML PhD life.