Demysifying Deep Neural Networks Through Interpretation: A Survey