Machine learning exercices

Introduction

The exercises proposed during the practical sessions of the machine learning course are based on the mini-courses available on the kaggle platform.

Kaggle is an online community of data scientists and machine learning practitioners. Kaggle allows users to find and publish data sets, explore and build models, work with other data scientists and machine learning engineers, and enter competitions to solve data science challenges.

Kaggle has embedded notebooks similar to jupyter notebooks that allows one to experiment and build machine learning algorithms in a fully web-based environment.

Exercices

We will ask you to fulfill the following mini-courses available in the "courses" section of kaggle:

• Intro to machine learning

• Pandas

• Intermediate machine learning

• Feature engineering

• Intro to deep learning

• Computer vision

Project

You will have to analyse and train a model on the following task:

Xente fraud detection challenge

At the end of your project, you should produce a report that contains: all the analysis of the dataset and the feature engineering that you did, the performance of your model, and a comparison with different other models. The quality of the scientific reasoning is equally as important as the performance of your model.

Assessment

You will be assessed of two works:

• A short written interogation that will be held at the end of the last lab session. The content will be based on the technical content of the kaggle exercices.

• The project: report and code should be uploaded to claco.