KernelMethods_DataChallenge

Data Challenge from Kaggle for the course Kernel Methods for Machine Learning from the M2-MVA (ENS-Paris Saclay).

Authors:

• Carlos Cuevas Villarmin
• Javier Alejandro Lopetegui González

Project Overview:

This repository contains our implementations for the 2024 Data Challenge for Kernel Methods course at ENS-Paris Saclay.

The challenge consists in an image classification task using CIFAR-10 dataset using kernels methods approaches.

Kernels implemented(kernels.py):

• Linear Kernel
• Polynomial kernel
• RBF Kernel
• Laplacian RBF Kernel

Feature extractor approaches(utils.py):

• Histogram of gradients
• SIFT
• Daysi

For the feature extractors we used the python package scikit-image.

Classifier implemented for the taks (classifiers.py):

• Kernel SVC One vs All (MulticlassKernelSVC)
• Kernel SVC One vs One (OneVsOneKernelSVC)
• Multivariate Kernel Ridge Classifier

Running the start.py file:

The file start.py contains the code to run a complete pipeline for the classification task. Particularly, it is configured for running by default the code for the las submission we made during the challenge with a public score of 0.644, the 4-th among all the participants.

To run the start.py file, follow these steps:

1. Make sure you have Python installed on your system.
2. Open a terminal or command prompt.
3. Navigate to the project directory
4. Run the following command: python start.py.
5. The application will start running and you will see the output in the terminal.

Note: Make sure you have all the necessary dependencies installed before running the start.py file. You can install the dependencies by running pip install -r requirements.txt in the project directory.