ETD system

Electronic theses and dissertations repository

 

Tesi etd-07172007-152325


Thesis type
Tesi di dottorato di ricerca
Author
Baronti, Flavio
email address
baronti@di.unipi.it
URN
etd-07172007-152325
Title
Hypothesis Testing with Classifier Systems
Settore scientifico disciplinare
INF/01
Corso di studi
INFORMATICA
Commissione
Relatore Starita, Antonina
Parole chiave
  • Statistical hypothesis testing
  • Machine Learning
  • Learning Classifier Systems
  • Decision Trees
Data inizio appello
21/06/2007;
Consultabilità
completa
Riassunto analitico
This thesis presents a new ML algorithm, HCS, taking<br>inspiration from Learning Classifier Systems, Decision Trees and<br>Statistical Hypothesis Testing, aimed at providing clearly<br>understandable models of medical datasets. Analysis of medical<br>datasets has some specific requirements not always fulfilled by<br>standard Machine Learning methods. In particular, heterogeneous<br>and missing data must be tolerated, the results should be easily<br>interpretable. Moreover, often the combination of two or more<br>attributes leads to non-linear effects not detectable for each<br>attribute on its own. Although it has been designed specifically<br>for medical datasets, HCS can be applied to a broad range of<br>data types, making it suitable for many domains. We describe the<br>details of the algorithm, and test its effectiveness on five<br>real-world datasets.<br><br>
File