a model trained on ecg data gathered from 350+ patients, which could predict different types of arrhythmia with the accuracy of 75%.
## Introduction ECG(electro cardio gram) holds a dominant place in the medical field for diagnosing different heart abnormalities. But still sometimes ecg reports are misdiagnosed due to presence of human error on the cardiologist side. The use of machine learning in finding relationship between the features of ecg data could drastically improve the quality of results in this diagnosis procedure. ## Working In this experiment a model is trained on ecg data taken from [UCI arrhythmia dataset] which contains 350+ patients data , which could predict 16 different types of arrhythmia with the accuracy of 75%. These results could be improved if substantial amount of data is available. In this experiment **Multi-class logistic regression** was used which is a classification method that generalizes logistic regression to multiclass problems, i.e. with more than two possible discrete outcomes. : https://archive.ics.uci.edu/ml/datasets/arrhythmia