Topic 24: Linear Algebra Q&A
🔥 For interviews, read these first:
LINEAR_ALGEBRA_DEEP_DIVE.md— frontier-lab deep dive: rank, eigendecomposition, SVD (with Eckart-Young), positive (semi)definiteness, matrix calculus (OLS gradient + Hessian), conditioning, projections.INTERVIEW_GRILL.md— 60 active-recall questions.
What You'll Learn
This topic covers the linear algebra you actually need for ML interviews:
- Rank, four fundamental subspaces, rank-nullity
- Eigendecomposition and the spectral theorem
- SVD as the universal factorization
- Positive (semi)definiteness — covariance, Hessians, kernel matrices
- Matrix calculus — derivatives that show up in OLS, ridge, neural nets
- Condition number and why it matters for optimization
- Projections and the geometric view of OLS
Why This Matters
Almost every ML algorithm is linear algebra at scale. PCA = eigendecomposition of covariance. Ridge = solving a regularized linear system. Neural networks = stacked linear maps with nonlinearities. SVD shows up in compression, recommender systems, low-rank adaptation (LoRA), embedding spaces.
Senior interviews probe whether you understand the operations — not just the names.
Next Steps
- Topic 37: MLE and MAP — links squared loss to Gaussian MLE, ridge to Gaussian MAP
- Topic 21: Dimensionality reduction — direct SVD application
- Topic 35: Kernel functions — Gram matrices and PSD