LADE Meeting

Emanuele Panizon

Title: Memory and control: Bits and bobs of understanding.

Abstract

Reinforcement Learning deals with how to act optimally to perform a predetermined task. Often, however, the agent lacks the complete information of the environment it is trying to manipulate: The instantaneous sensorial inputs, e.g., may not be sufficient to have an optimal (or even half-decent) performance. In this case the task may require storing past observations. But how to learn what should be retained in memory, and what should be quickly forgotten? Using the task of olfactory search, common in nature to animals from insects to rodents, I will try to give a general sense of how the algorithmic construction of memory may be included and exploited in the reinforcement learning framework.

Date
Mar 26, 2024 11:30 AM — 12:30 PM
Event
LADE Meeting
Location
Sala Riunioni RIT, Area Science Park
Località Padriciano 99, Trieste, 34149
Area Science Park - RIT
Area Science Park - RIT
Research Institute

The Institute of Research and Innovation Technology (RIT) at Area Science Park carries out cutting-edge research and provides services and consulting to public and private-sector users through its three laboratories equipped with state-of-the-art technology.