mathfoundations
Linear algebra
Vectors, matrices, transformations and decompositions — the mathematical backbone of ML.
Уровни глубины
L0Intro~4ч
Knows what a vector and a matrix are; can multiply a matrix by a vector with pencil and paper.
L1Basics~20ч
Can compute eigenvalues, transpose, dot products, matrix inverse; understands linear independence and span.
L2Working~30ч
Applies SVD, PCA, LU/QR decompositions from scratch; comfortable with NumPy/torch tensor ops.
L3Advanced~40ч
Analyses conditioning, numerical stability; uses Kronecker products, matrix calculus for gradient derivation.
L4Research~80ч
Contributes to randomised linear algebra, sketching, or low-rank approximation research.
Ресурсы
L0 — Intro
L1 — Basics
L2 — Working
Ведёт к
- PrerequisiteNeural networksL1→L2
- PrerequisiteBackpropagationL2→L2
- RelatedOptimization
- RelatedProbability theory
- PrerequisiteNumerical methodsL1→L1
- PrerequisiteConvex optimizationL2→L1
- PrerequisiteLinear regressionL1→L1
- PrerequisitePCA and dimensionality reductionL2→L1
- RelatedEmbeddings and representation learning
- PrerequisiteGraph neural networksL1→L1