HeartDisease - EDA,FE and ML Prediction

Project Description:

Used Heart Disease UCI dataset to analyse different parameters which often led to a Heart Attack
Performed Univariate, Bivariate, Multivariate Analysis of different parameters; used EDA and Advanced EDA (Pandas Profiling, Autoviz, Pandas Describe) to graphically represent the data and deduce certain key observations from it
Created Correlation matrix on Categorical Data using Pearson and Spearman Method and also found the feature importance of various parameters which gave useful insights into the data
Used LabelEncoder to encode the data
Performed Feature Selection on the Data using 5 different Feature Selection models - LinearSVC, Lasso, Chi-2, Regressive Elimination - Logistic Regression and VarianceThreshold
For more accuracy, I considered a combination of 2 features as new features to increase the accuracy of the feature selection and prediction
Used 4 ML models for prediction - Linear Regression, Random Forest classifier, Logistic Regression and XGBoost, did Hyperparameter tuning of the most accurate model to further increase accuracy

Name		Name	Last commit message	Last commit date
Latest commit History 21 Commits
Heart Disease - EDA, FE and ML Prediction.ipynb		Heart Disease - EDA, FE and ML Prediction.ipynb
README.md		README.md
heart.csv		heart.csv