sakoya00 / diabetes_ml

Predicted diabetes progression after one year using seven different regression ML techniques on the LARS diabetes dataset

diabetes_ml

Predicted diabetes progression after 1 year using seven different regression ML techniques and the predictors (age, sex, BMI, etc) found in the LARS diabetes dataset. Selected best model of the seven (LASSO) using RMSE and R^2 on test data.

Required Packages/Modules

• pandas
• numpy
• sklearn
• matplotlib
• math

Diabetes Predictors

• age in years
• sex
• BMI
• average blood pressure
• s1 tc, T-Cells (a type of white blood cells)
• s2 ldl, low-density lipoproteins
• s3 hdl, high-density lipoproteins
• s4 tch, thyroid stimulating hormone
• s5 ltg, lamotrigine
• s6 glu, blood sugar level

Regression Techniques Used

• ElasticNet
• LASSO
• Ridge
• Linear Regression
• Polynomial Regression
• Support Vector Regression (SVR)
• Random Forest Regression