Computational Bioacoustics

7 Audio recognizers

Introduction

The mathematical description of the probabilistic modelling was formulated by Thomas Bayes (1763), who laid the foundations of the statistical decision-making theory. In brief, Bayes’ theory is grounded on two assumptions: (i) that the decision problem can be specified in probabilistic terms and (ii) that all relevant probabilities are known or can be derived from the observed data and/or some prior knowledge about the problem. These two assumptions are in common use in the contemporary philosophy of statistical classification. Bayes’ theory has laid the groundwork for the elaboration of a great number of classification methods.

7.1 Overview of classification methods

The implementation of the various species ...