SigmaAnalytics & Tech
B5PythonIntermediateAbout 130 min to complete

Linear Algebra

The mathematics underneath every regression, PCA, and neural network — implemented from scratch, then done properly.

Notebook

About 130 minutes to complete

Enrol to read the notebook and download it to run.

What this module covers

The mathematics underneath every regression, PCA, and neural network — implemented from scratch, then done properly.

By the end of it

  • Distinguish element-wise multiplication (*) from matrix multiplication (@)
  • Compute dot products, matrix products, and understand the shape rules that govern them
  • Interpret the dot product geometrically as projection and similarity
  • Use transpose, identity, inverse, determinant, trace, and rank correctly
  • Solve linear systems with np.linalg.solve rather than inverting matrices
  • Recognise ill-conditioned problems via the condition number
  • Fit linear regression three ways and explain the numerical trade-offs
  • Compute eigenvalues and use SVD for dimensionality reduction