The electrocardiogram (ECG) signal is used to diagnose various Cardiac ailments as it holds the fundamental information to make appropriate decisions about different types of heart diseases. Hence several strategies were proposed to extract critical features from the ECG signal with highest accuracy which helps for the autonomous detection of Cardiac ailments. A methodology has been proposed in this work for state of the art in automatic detection of Cardiac ailments which include pre-processing, Feature extraction and Classification steps. A Butterworth third order band pass filter is used in pre-processing step and a four level Maximal overlap discrete wavelet packet transform (MODWPT) with symlet as mother wavelet is used for feature extraction step. Finally, for classification of considered three Cardiac ailments from MIT-BIH database i.e., Arrhythmia, Congestive Heart Failure and Atrial Fibrillation from Normal Sinus rhythm, five supervised Machine learning algorithms i.e., Support vector machine (SVM), K-nearest neighbour (KNN), Naive Bayes (NB), Decision tree (DT) and Random Forest (RF) were used which gives an overall accuracy of 90.83%, 90.56%, 90.28%, 91.39% and 91.94% for each classifier respectively. Clearly, random forest classifier for the proposed methodology gives better accuracy of the model for multiclass classification of cardiac ailments.