Raga Analysis Using Entropy

Abstract

If P(E) is the probability of an event, the information content of the event E is defined as I(E) = −log2(P(E)). Events with lower probability will signal higher information content when they occur. The probability of a raga note, and hence its information content, depends on the raga concerned. The important raga notes will obviously be having higher probabilities. On the other hand, a weak note in a raga cannot be thrown away either for it would be carrying more surprise! The strength of entropy analysis lies here (entropy is the mean information content of a random variable).