Let's say I wanna classify the audio files of race car exhaust and normal car exhaust. I attain few features such as ZCR, energy and enthropy of energy from the training set and I wanna build the prediction model out of those. How do I do that? Is there some machine learning model that I could use?
This is a machine learning problem, and it can be classified using any good classification algorithm like 1) Linear Classifiers: Logistic Regression, 2) Naive Bayes Classifier, 3) Nearest Neighbor, 4) Support Vector Machines, 5) Decision Trees, 6) Boosted Trees, 7) Random Forest, 8) Neural Networks. Matlab has in-built classifiers, and you can train your model very easily.
As you have already found out the features you should get some good classification results using any of these classifiers. All of the features that you mentioned are time domain-based. Also, you can incorporate some frequency-based features like 1) Spectral Centroid and Spread (C.G and deviation from CG), 2) Spectral Entropy, 3) Spectral Flux (Measure of Spectral Change over successive frames), 4) Spectral Roll-off (Spectral shape descriptor of an audio signal) which may increase the classification accuracy.
For more audio signal features you can go through this book "Giannakopoulos, Theodoros, and Aggelos Pikrakis. Introduction to audio analysis: a MATLAB® approach. Academic Press, 2014."