ENTROPIAS DE SHANNON E RÉNYI APLICADAS AO RECONHECIMENTO DE PADRÕES

This paper presents and discusses the concepts of e ntropy and mutual information in Information Theory applied to pattern recognize, based mainly on the i deas of Shannon and Renyi. Based on the diversity o f application of the theory, as in engineering, stati stics, economics and informatics and the few litera ture in Portuguese, is exposed a way of using the entropy a pplied to the academic researches, with an applicat ion showed for the construction of decision tree. Are a lso discussed similarities and differences with the physical