DIABETES RISK CLASSIFICATION

Machine Learning / Research

Diabetes risk classification on PIMA dataset


WHAT IS IT?

A machine learning research project focused on diabetes risk classification using the PIMA Indians Diabetes dataset.

NOTE

This entry documents applied machine learning research work.

DETAILS

  • > Replicated published research benchmarks
  • > Used GridSearchCV to tune Random Forest
  • > Developed a PCA hybrid framework
  • > Combined dimensionality reduction with Random Forest and XGBoost
  • > Evaluated whether selected clinical features could retain predictive performance after dimensionality reduction

TECHNOLOGIES

Python · Scikit-learn · PCA · Random Forest · XGBoost · GridSearchCV

RESULTS

  • > 78.62% accuracy compared with the paper's 77%
< BACK TO PROJECTS