ETD

Archivio digitale delle tesi discusse presso l'Università di Pisa

Tesi etd-09182019-150521


Tipo di tesi
Tesi di laurea magistrale
Autore
ROSSETTO, FEDERICO
URN
etd-09182019-150521
Titolo
Memory Augmented Neural Networks in Reinforcement Learning
Dipartimento
INFORMATICA
Corso di studi
INFORMATICA
Relatori
relatore Prof. Bacciu, Davide
Parole chiave
  • neural networks
  • memory augmented neural networks
  • reinforcement learning
Data inizio appello
04/10/2019
Consultabilità
Completa
Riassunto
This work explores the capabilities of the current Reinforcement Learning algorithms and the Memory Augmented Architectures in tackling tasks that requires the use of external memory. It also proposes a new architecture that improves the model capabilities in solving this type of tasks. The models have been tested on two different grid-based maze tasks, where the agent needs to remember some information to act accordingly later. We compare our proposed model with Neural Turing Machines and Long Short Term Memory in their capabilities of solving this task.
File