Skip to content

mciaravino/lipschitz-gradient-descent

Repository files navigation

Lipschitz Constant Estimation via Gradient Descent

A MATLAB implementation that uses gradient descent to estimate the Lipschitz constant of a function, with graph construction and geometric slit visualization.

Files

  • drivingScript.m — Main script to run the project
  • gd_compute_lip.m — Computes global Lipschitz constant of the filling map over all graph edges
  • gd_local_lip.m — Computes worst Lipschitz ratio over edges incident to a single node (used inside descent loop)
  • gd_dlookup.m / gd_get_row.m — Distance cache utilities
  • hl_build_graph_from_points.m — Constructs graph from point data
  • hl_draw_slits_iter.m — Iterative slit drawing
  • hl_overlay_grid_with_slit_holes.m — Grid and slit overlay visualization
  • hl_plot_graph_edges.m — Graph edge plotting

Presentation

Slides in progress — view draft

Requirements

MATLAB (R2021a or later recommended)

About

MATLAB implementation of gradient descent to estimate the Lipschitz constant, with graph construction and slit-based geometric visualization.

Topics

Resources

Stars

Watchers

Forks

Contributors

Languages