Innovating instruction of communication theory with machine learning and speech analysis

Abstract

Foundational work of Claude Shannon in 1950's provided direction for theory of communication for many decades thereafter. Emphasis of research was on development of new coding and decoding techniques in order to make effective use of all available bandwidth. With the advent of Internet, and especially of Internet commerce, security aspects of communication theory became much more important. As research in these questions enters a mature stage, the challenge for teachers of communication theory is to find novel ideas, new applications and new interdisciplinary topics in order to engage students as well as to prepare them to tackle challenges they will encounter in their engineering careers later on. In our paper we propose to apply to teaching emerging connections of information theory with two other scientific disciplines: machine learning and analysis of speech.