Heart Attack Prediction by Machine Learning with Source Code and Documentation

Project Discription

Heart attacks are becoming more prevalent in people's daily lives as a direct result of the general improvement in their living conditions. Detecting and diagnosing this diabetes condition in its early stages is currently a pressing need in the modern world. A crucial factor in the categorization of diabetes disease is the process of diagnosing diabetes and analysing data related to the disease. It is necessary to construct a classifier that is not only valid but also easy on the wallet and the convenience factor. This work has developed a model that predicts the diabetic situation of the patient as per different observation features. Detection of disease in this work is termed as SFLANN and it have three module first is selection of features from available set of features than second is training of neural network was performed from available set of filtered features. Finally in third module testing was performed on the trained neural network. Here selection of features was done by Shuffled Frog Leaping Algorithm and training of Error Back Propagation Neural Network was performed. Hence objective of this work was to reduce number of features with increase  detection accuracy.  Experiment was done on real dataset of diabetic paitents. Result shows that proposed model has improved the average accuracy value by 26.43%. Use of error back propagation neural network has improved the testing performance. improved the f-measure parameter value by 16.14% as compared to previous work proposed.

Project Sample Image