Heart-Failure-Prediction-with-Machine-Learning

Creating ML model using "Heart Failure Prediction Dataset" (from kaggle.com) to predict Heart failure of patients.

The data is downloaded from https://www.kaggle.com/datasets/fedesoriano/heart-failure-prediction

Here I created two different model that both can predict heart failure with 0.88 F1-score.

The steps:

• Cleaning the data
• Exploratory data analysis
• Scale the features (StandardScaler)
• Model selection, tested models were:
  • RandomForest
  • Logistic Regression
  • Support Vector
  • KNeighbors
  • xgboost
  • Gaussian Naive Bytes Classifier
• Hyperparameter tuning for Support Vector classifier.
• Build another model with Deep learning.
• Identify the importance of the features using SHAP.