Skip to content

Latest commit

 

History

History
32 lines (26 loc) · 2.01 KB

README.md

File metadata and controls

32 lines (26 loc) · 2.01 KB

AP_PROG Machine Learning Assignment

Classification task keywords :

scRNAseq data-based tissue prediction , Highly unbalanced, Very High dimensionality

Regression task keywords :

Life expectancy prediction, Model explainability, Error Analysis

Website of this reposiory

Created for the evaluated assignment of the AP_PROG Machine Learning course at AgroParisTech given by Mr. Vincent GUIGUE. The details of the asisgnment are given here (in french).

In the present repository you will find the html rendered versions of my notebooks for the Classification task I decided to investigate and another for the Regression task I chose. I also made available the actual notebooks for anyone wanting to play with the data and further investigate the problem themselves.
You will need to download specified datasets by yourself as they are too heavy to be shared on github.

Note that these notebooks take quit some time to run and they produce big intermediary files so you may not want to run them if you do not have time / sufficient memory available.

These notebooks were made using :

  • numpy 1.26.4
  • pandas 2.2.2
  • random
  • pickle
  • matplotlib 3.9.2
  • seaborn 0.13.2
  • sklearn 1.5.2
  • umap-learn 0.5.6
  • optuna 4.0.0
  • shap 0.45.1
  • imbalanced-learn 0.12.4

Classification task : Predicting tissue of extraction of B cells.

The goal is to use single-cell RNAseq profiles to predict the tissue from which each B cell was extracted. In the dataset I chose, we have 9 heavily unbalanced classes which we try to predict. I found the dataset on the awesome Tabula Sapiens Database.

Regression task : Predicting Life expectancy.

For this problem, the task was to be able to accurately predict Life expectancy based on health and economy factors. The dataset can be found here on Kaggle.