/02620_Cancer_Classification_Project

Primary LanguageJupyter Notebook

Course Project of Group 1 - Cancer classification using machine learning

This is the course project for 02-620: Machine Learning For Scientist in Spring 2022. And oour project is about cancer classification using machine learning.

Group members

  • Yuxuan Wu
  • Eric Li
  • Yifan Wu
  • Xin Wang

Dependencies (only for model implementation section)

  • Python3
  • scikit-learn
  • pandas
  • numpy
  • pytorch

All jupyter notebooks are self-contained and runnable!

Original Datasets (too large to include)

BRCA

LUAD

Data preprocessing and DGE

We used R to perform Differential Gene Expression (DGE), which is one of our feature selection methods. And the other is selectKBest function in Python.

R scripts are under ./Preprocessing_and_DGE

Models

Self Implemented Machine Learning Algorithm (Decision Tree and AdaBoost)

self-implemented decision tree model: ./Model/Decision_tree/decision_tree.ipynb
self-implemented adaboost model: ./Model/AdaBoost/adaboost.ipynb

MLP with Pytorch

./Model/MLP

Follow-Up Evaluation and Results

Model evaluation scripts

./Performance

Model evaluation results

./Results/Performance

Feature selection scripts

./Feature_importance

Feature selection results

./Results/Feature_importance

Additional plots that didn't contain in the report because of page limit

./Additional_plots