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.
- pandas
- numpy
- sklearn
- matplotlib
- math
- 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
- ElasticNet
- LASSO
- Ridge
- Linear Regression
- Polynomial Regression
- Support Vector Regression (SVR)
- Random Forest Regression