In Carnatic music, taala plays a vital role. Taala classification in Carnatic music is an area in which very few works have been reported so far. For any audio sample, taala is the most unique factor. Thus, it is important to study the variations of taala in an audio to know which category it belongs to. In this paper, we aimed at classifying four major taalas namely Aadi, Rupaka, MishraChapu and KhandaChapu in Carnatic music by using various machine learning algorithms and Deep Neural Networks (DNNs). We found out the correlation between the attributes and identified the best attribute that contribute well to the classification of taala from the dataset. We implemented the algorithms such as Random forest, K- Nearest Neighbour, Principal Component Analysis (PCA) with Logistic Regression (LR) and Support Vector Machine (SVM) with kernel Gaussian Radial Basis Function(RBF) and observed that PCA with LR and SVM with RBF performed the best classification with 98.11% and 98.1% accuracies, respectively. Also, we worked with DNN incorporating six hidden layers and resulted in a better performance of 97.50% to classify the taalas.