Diagnostic Aid for Heart Disease Using Data

Seed Project for Medical Platform

Algorithms Comparison

• Best algorithm to predict hearth disease: SVM

Discovery

• Algorithms with best accuracy

Algorithm	Data Type	Scale	Accuracy
SVM (kernel = rbf)	Feature Scaled Data	0, 1	82 %
SVM (kernel = linear)	Nominal Data	0, 1, 2, 3, 4	77 %

Expectation

• Predicted heart disease using 13 features

• Apply similar logic to other disease

• If these algorithms are applied to all disease, computer can diagnose probability of each disease with given features.

• Doctors will make better diagnosis with help of these algorithms applied to medical data.

In the Future

• GitHub organization: MedicalDataAnalysis

• repository: Heart_Disease-Wisconsin-

• Plan to make repositories for other disease whenever we get the data

• Build medical platform using models we created

Learn More

• The report of this project is in _Report folder.

• If you want to contribute to this repository, feel free to contact us!