Case Study 14

Gradient Descent from Scratch

Univariate & multivariate optimizers with a convergence study

2025
Code
PythonNumPySciPyMatplotlib
Back to projects
Key impact
Implemented univariate and multivariate gradient-descent optimizers from scratch, including the cost functions and their analytic gradients.
GRADIENT DESCENTθ₀θ*η sweep · convergence study
Representative mockup

What I did

3
  1. 01

    Implemented univariate and multivariate gradient-descent optimizers from scratch, including the cost functions and their analytic gradients.

  2. 02

    Ran convergence experiments across learning rates and visualized loss surfaces and descent trajectories.

  3. 03

    Unit-tested the optimizer updates against numerical references to confirm correctness.

Tech stack

PythonNumPySciPyMatplotlib
More projects
← All projects