Skip to content

README.md

Latest commit

 

History

History
81 lines (52 loc) · 2.51 KB

README.md

File metadata and controls

81 lines (52 loc) · 2.51 KB

Disease Prediction System using Machine Learning

Nexus Info Project-3 (AI/ML domain)

Made By: Kumar Shantanu (4th Year CSE Student at KIIT University, Bhubaneswar, Odisha, India)

My Email ID: shantanunitw01@gmail.com

Overview

This project focuses on predicting heart disease using machine learning algorithms and evaluates their performance on a dataset. The goal is to identify the most effective model for accurate heart disease prediction.

Problem Statement

Heart disease is a significant health concern globally, and accurate prediction models are crucial for timely intervention. This study aims to enhance the reliability of heart disease prediction by comparing and fine-tuning machine learning algorithms.

Project Description

Heart disease prediction is addressed through the application of various machine-learning algorithms on a comprehensive dataset. The project aims to discover patterns in the data to improve the accuracy of predicting heart disease.

Project Objectives

  1. Compare ML algorithm performance for heart disease prediction.
  2. Identify the most effective ML approach for accurate prediction.
  3. Fine-tune models for enhanced accuracy.

Problem-Solving Approach

  1. Understanding the Problem
  2. Data Collection
  3. Data Preprocessing
  4. Exploratory Data Analysis (EDA)
  5. Model Selection
  6. Best Model Selection
  7. Hyperparameter Tuning
  8. Model Training
  9. Model Testing
  10. Performance Metrics
  11. Model Interpretability
  12. Validation and Testing
  13. Visualization of Results

Results

Technology Used

  • Programming Language: Python
  • ML Libraries: Scikit-learn, SHAP
  • Data Analysis: Pandas, NumPy, Matplotlib, Seaborn
  • Metrics: Accuracy Score, Precision, Recall, F1 Score, SHAP values, Confusion Matrix
  • Tools: GridSearchCV
  • Visualization: Seaborn, Matplotlib, SHAP

Modelling

ML models:

  • Logistic Regression
  • Decision Tree
  • Random Forest
  • Support Vector Machine

Conclusion

Model Performance:

  • Random Forest achieved the highest accuracy, 85.37%, making it the best-performing model.
  • Other models showed competitive accuracies ranging from 73.17% to 82.93%.

Hyperparameter Tuning:

  • Random Forest was fine-tuned with optimal hyperparameters, enhancing its accuracy.

Model Comparison and Selection:

  • Random Forest's superior performance makes it the preferred choice for heart disease prediction.

Validation and Testing:

  • Models were extensively tested and validated to ensure the accuracy, reliability, and robustness of the heart disease prediction system.