Mathematics & computational learning
Lessons that explain the mathematics used in real projects, from concept and equations to worked examples and implementation.
Linear algebra, calculus, probability, statistics, and optimization with a focus on where the mathematics appears in AI.
6 items
Problem modeling, objectives, constraints, evolutionary methods, and metaheuristics for decision and search problems.
3 items
Connecting code, mathematical models, numerical experiments, data, and reproducible implementation.
4 items
Applied mathematics behind system behavior, rates, scoring, and engineering algorithms.
1 items
Search by concept or title, then narrow the results by category and level.
14 items
Convexity, positive semidefiniteness, and the role of the Hessian in convex QP.
A mathematics lesson connecting convex structure to the Convex QP Solver.
Vector geometry, similarity, and its role in search and AI.
A mathematics lesson connecting linear algebra to representation and relevance.
How a decision problem becomes variables, an objective, and constraints.
A foundational lesson on translating a decision problem into variables, an objective, and constraints.
From Lagrangians and KKT conditions to barrier methods and Convex QP.
A numerical-computing lesson connecting Lagrangians, KKT conditions, barrier methods, and Convex QP.
From data representation and linear transforms to matrices inside model layers.
A mathematics lesson connecting vectors and matrices to data representation and AI model layers.
From one-variable derivatives to multivariable gradients and gradient descent.
A lesson connecting derivatives and gradients to loss functions and gradient-based learning.
The core ideas and mathematics behind approximate search with GA and DE.
A lesson introducing population-based search, Genetic Algorithm, Differential Evolution, and the exploration/exploitation trade-off.
From equation residuals to independent checks of numerical solutions.
A scientific-computing lesson connecting Python implementation with residuals, numerical answers, and independent checks.
From matrix products and scores to softmax and weighted Value aggregation.
A mathematical explanation of attention using Q, K, V, and the core equations.
How limits and time windows become rates, intervals, and burst tolerance.
The mathematical ideas behind Token Bucket and GCRA, connected directly to RateLimitEngine.
From residuals and norms to condition numbers and numerical sensitivity.
A scientific-computing lesson about error, sensitivity, and validation.
Probability and inference foundations for statistical and machine-learning models.
A mathematics lesson about uncertainty, likelihood, and Bayes in AI.
A guided explanation of the 2017 Transformer paper, from the sequence-transduction problem to scaled dot-product attention and the encoder-decoder architecture.
A paper explainer that keeps the original source authoritative while unpacking its main ideas, equations, and engineering implications.
A guided reading of a compact autograd engine and neural-network library, using it to make reverse-mode differentiation concrete.
A project explainer showing how a small public codebase can make computational graphs, local derivatives, and backpropagation easier to inspect.
This area is for conceptual, mathematical, and technical lessons, with real connections to projects, research, and solutions.
