Tesi etd-09172018-093632 |
Link copiato negli appunti
Tipo di tesi
Tesi di laurea magistrale
Autore
SETZU, MATTIA
URN
etd-09172018-093632
Titolo
Building global interpretable classifiers via local explanations
Dipartimento
INFORMATICA
Corso di studi
INFORMATICA
Relatori
relatore Guidotti, Riccardo
relatore Monreale, Anna
relatore Turini, Franco
relatore Monreale, Anna
relatore Turini, Franco
Parole chiave
- AI
- Data Mining
- XAI
Data inizio appello
05/10/2018
Consultabilità
Non consultabile
Data di rilascio
05/10/2088
Riassunto
In our thesis we formally define the "Local to Global" (LtG) problem, review its literature, and propose three frameworks to address it. Informally, the LtG is defined as the creation of crystal boxes through combination of several explainable models.
Our contributions consists of a meta-logic semantic to manipulate and transform crystal boxes, and three LtG frameworks: i) a pure scoring framework, a meta-logic based framework and a hierarchical clustering framework.
Our contributions consists of a meta-logic semantic to manipulate and transform crystal boxes, and three LtG frameworks: i) a pure scoring framework, a meta-logic based framework and a hierarchical clustering framework.
File
| Nome file | Dimensione |
|---|---|
Tesi non consultabile. |
|