Adeptness Associative Learning Method for Real-Time Cardiac Arrhythmia Detection

It is vital for the automated system to accurately detect and classify ECG signals very fast to provide a useful means for tracing the heart’s health in the real time. Making a random training set can lead or cause negative results. Simply, training set must designed carefully to consider all possible classes of the overall arrhythmia, so as to train the algorithm with the right group. Not only, considering all possible arrhythmia, but also with the same ratio, which means giving the algorithm a richness group to be trained in the right way with effective training set. Therefore, a means is needed for determining which record in the main file to be selected to replace the removed one from the active training data set. Our proposed methodology automatically trains the classifier model, using efficient set. Accordingly the experimental works show an improvement in the performance of some classification models to detect cardiac arrhythmia.