https://github.com/khushiisaxena/heart-failure-prediction
A predictive model for heart failure using machine learning algorithms. This research project focuses on enhancing the accuracy of heart failure prediction models through hyperparameter optimization and feature selection. We analyze and enhance eight supervised learning classification algorithms to achieve high prediction accuracy.
https://github.com/khushiisaxena/heart-failure-prediction
Last synced: about 1 month ago
JSON representation
A predictive model for heart failure using machine learning algorithms. This research project focuses on enhancing the accuracy of heart failure prediction models through hyperparameter optimization and feature selection. We analyze and enhance eight supervised learning classification algorithms to achieve high prediction accuracy.
- Host: GitHub
- URL: https://github.com/khushiisaxena/heart-failure-prediction
- Owner: khushiisaxena
- Created: 2024-04-06T17:50:30.000Z (11 months ago)
- Default Branch: main
- Last Pushed: 2024-08-07T14:00:48.000Z (7 months ago)
- Last Synced: 2024-11-20T10:51:01.886Z (3 months ago)
- Language: Jupyter Notebook
- Homepage:
- Size: 2.98 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Heart-Failure-Prediction
This research project aims to improve the accuracy of heart failure prediction models using hyperparameter optimization and feature selection. 8 supervised learning classification algorithms namely Logistic Regression, KNN, Naive Bayes, SVM, Decision Tree, Random Forest, LightGBM and XGBoost predictions were analyzed and improved upon using hyperparameter optimization and feature selection. An accuracy of 87.83% was achieved using Random Forest for all 12 features and KNN for top 5 features (selected after feature selection).
## Algorithms Used
The following supervised learning classification algorithms were analyzed:
- Logistic Regression
- K-Nearest Neighbors (KNN)
- Naive Bayes
- Support Vector Machine (SVM)
- Decision Tree
- Random Forest
- LightGBM
- XGBoost
## Feature Selection
Feature selection techniques were applied to identify the most significant features contributing to the prediction of heart failure. The top 5 features were selected based on their importance scores.
## Hyperparameter Optimization
Hyperparameter optimization was performed to enhance the performance of the machine learning models. Grid search and random search methods were used to find the optimal hyperparameters for each algorithm.
## Results
- Random Forest: Achieved an accuracy of 87.83% using all 12 features.
- KNN: Achieved an accuracy of 87.83% using the top 5 selected features.