ZetaDiseasePrediciton

Zeta Disease Prediction for Carrier Inc by Raviteja Neravati

THE CHALLENGE: Zeta Disease Prediction

INTRODUCTION

Mars Mission Control needs a good data-driven system for predicting Zeta Disease infection on the International Mars Colony. Used the _zeta-disease_training-data_ dataset to build a model that can predict who will be infected by Zeta Disease. Applied multiple ML models to the _zeta-disease_prediction-data_ dataset to predict who will be infected by Zeta Disease and choosed the best one based on the highest score.

DATASET

The dataset includes 9 columns with information on 800 people.

1. age : in years
2. weight : body weight in pounds (lbs)
3. bmi : Body Mass Index (weight in kg/(height in m)2)
4. blood_pressure : resting blood pressure (mm Hg)
5. insulin_test : inuslin test value
6. liver_stress_test : liver_stress_test value
7. cardio_stress_test : cardio_stress_test value
8. years_smoking : number of years of smoking
9. zeta_disease : 1 = yes; 0 = no

ML MODEL

1. Descriptive Study - looking at the Histogram of the columns
2. Standardised the continous variables
3. Splitted the data in to 80/20 ratio
4. Trained KNN, SVM, DT, and RF classification models and performed model evaluation on test data
5. choosen the best model (SVM - high score) and predicted the Zeta_Disease (1 = yes, 0 = No)