https://github.com/pronzzz/diabetes-prediction

Diabetes prediction using a KNN model and Pima Indian Diabetes Dataset
https://github.com/pronzzz/diabetes-prediction

Last synced: 14 days ago
JSON representation

Diabetes prediction using a KNN model and Pima Indian Diabetes Dataset

• Host: GitHub
• URL: https://github.com/pronzzz/diabetes-prediction
• Owner: pronzzz
• Created: 2024-01-23T01:23:08.000Z (10 months ago)
• Default Branch: main
• Last Pushed: 2024-01-28T04:35:06.000Z (10 months ago)
• Last Synced: 2024-01-28T05:21:44.084Z (10 months ago)
• Language: Jupyter Notebook
• Size: 4.65 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
## **🤖 Diabetes Prediction using Machine Learning 📊**

- Pandas, NumPy, Matplotlib, and Seaborn are used for data manipulation and visualization 📈

- Data Cleaning 🧹

  - Missing Values: No missing values 😊

  - Duplicated Values: None found 👍

- Exploratory Data Analysis 🔎

  - Outcome Distribution: Balanced ⚖️

  - Outlier Detection: A few in some features 👀

- Data Preprocessing 🛠️

  - Standard Scaling: Applied for better model performance 🎚️

  - Label Encoding: Outcome variable (0/1) 🎯

- Model Training and Evaluation 🚂

  - K-Nearest Neighbors Classifier 🤝

  - Grid Search for Optimal Hyperparameter (k) 🔧

  - Max Train Score: 94.23% at k = 14 🥇

  - Max Test Score: 88.89% at k = 13 🏆

- Performance Evaluation 📊

  - Confusion Matrix: Shows model's prediction accuracy ⬜⬛

  - Classification Report: Detailed metrics (Precision, Recall, F1-score) 👌

- Conclusion 🏁

  - KNN Classifier with k = 13 provides the best balance between train and test scores 🏆

  - Model can effectively predict diabetes based on patient's features 😄