Advanced Machine Learning Project

This repository contains the python files for our Advanced Machine Learning report. For this project we had to analyse two manifold learning algorithms: Locally Linear Embedding (LLE) and Hessian LLE (HLLE). There are 4 mains python notebooks:

• create_toy_dataset.ipynb: Creates the toy datasets used for qualitative results and comparison
• effect_hyper_parameters.ipynb: Analysis on the toy datasets
• silhouette.ipynb: Analysis on the Calltech101 silhouette dataset
• cancer_manifold.ipynb: Analysis on the Wisconsin Breast Cancer dataset.

For convenience, all 4 notebooks are merged into a single one: to_be_rendered/Final.ipynb.

Running the code

In addition to common python libraries (numpy, matplotlib, ...) this project has a few other dependencies mainly scikit. If you have trouble running this code, you can create a virtual python environment with the requirements.

Authors

• Thomas Havy
• Antoine Weber