Diabetes Prediction

Analyzing various attributes on which diabetes directly depends and predicting if a person with a given health conditions has diabetes.

Motivation

Project was created to make a comparison between the accuracy of predictions obtained from Support Vector Machine and Naivebayes algorithm. The goal of the project is to tune the two algorithms in such a way that the accuracy of prediction is maximized.

Data-set

Data-set being used in this project can be found on kaggle or you can use this link to go directly to the data-set.

Screen shots

-HeatMap

-Output using NaiveBayes

index	precision	recall	f1-score	support
0	0.82	0.85	0.83	150
1	0.70	0.64	0.67	81
micro avg	0.78	0.78	0.78	231
macro avg	0.76	0.75	0.75	231
weighted avg	0.78	0.78	0.78	231

-Output using SVM

index	precision	recall	f1-score	support
0	0.82	0.87	0.85	103
1	0.70	0.61	0.65	51
micro avg	0.79	0.79	0.79	154
macro avg	0.76	0.74	0.75	154
weighted avg	0.78	0.79	0.78	154

How to use?

Clone

• Clone this repo to your local machine using https://github.com/arnavgarg123/DiabetesDatabase-NaiveBayes-SVM.git

Setup

• Make surer you have jupyter notebook installed on your system with python 3 kernel.
• Using terminal/cmd navigate to the folder containing the files of this repo and run the command juputer-notebook.
• Now open NaiveBayes-Diabetes.ipynb for NaiveBayes and SVM-Diabetes.ipynb for SVM on jupyter notebook.

Contributing

Step 1

• Clone this repo to your local machine using https://github.com/arnavgarg123/DiabetesDatabase-NaiveBayes-SVM.git
  

Step 2

• HACK AWAY!
  

Step 3

• Create a new pull request
  

License

This project is licensed under the MIT License - see the LICENSE.md file for details