Szczegóły publikacji
Opis bibliograficzny
Motivated agent with semantic memory / Janusz Starzyk, Marcin Kowalik, Adrian HORZYK // W: ECAI 2023 : 26th European Conference on Artificial Intelligence : including 12th conference on Prestigious Applications of Intelligent Systems (PAIS 2023) : September 30 - October 4, 2023, Kraków, Poland : proceedings / ed. by Kobi Gal, [et al.] ; European Association for Artificial Intelligence (EurAI), Polish Artificial Intelligence Society (PSSI). — Amsterdam : IOS Press BV, cop. 2023. — (Frontiers in Artificial Intelligence and Applications ; ISSN 0922-6389 ; vol. 372). — ISBN: 978-1-64368-436-9; e-ISBN: 978-1-64368-437-6. — S. 2202-2209. — Bibliogr. s. 2208-2209, Abstr. — Dod. abstrakt dostępny w: https://ecai2023.eu/acceptedpapers [2023-11-06]
Autorzy (3)
- Starzyk Janusz A.
- Kowalik Marcin
- AGHHorzyk Adrian
Dane bibliometryczne
ID BaDAP | 149128 |
---|---|
Data dodania do BaDAP | 2023-11-06 |
Tekst źródłowy | URL |
DOI | 10.3233/FAIA230517 |
Rok publikacji | 2023 |
Typ publikacji | materiały konferencyjne (aut.) |
Otwarty dostęp | |
Creative Commons | |
Wydawca | IOS Press |
Konferencja | 26th European Conference on Artificial Intelligence |
Czasopismo/seria | Frontiers in Artificial Intelligence and Applications |
Abstract
This paper introduces a motivated agent scheme that enables an agent to create its own goals using prior knowledge about its environment. A motivated agent operates in a dynamically changing environment and is capable of setting and achieving its own goals, as well as those set by the designer. The agent has access to additional knowledge about the environment, which is represented in associative semantic memory. This memory is constructed based on ANAKG associative knowledge graphs, which have been shown to have several advantages over other semantic memories for processing symbolic sequential inputs. They are easy to organize and train. In this paper, we demonstrate that a motivated agent with semantic memory learns to achieve its goals more easily and utilizes environmental resources more effectively. To simplify comparisons with reinforcement learning, we represent the environment as an environmental graph that shows the principles governing it. By exploring the environment, the agent learns these principles and can use them to accomplish its tasks. Our experiments and tests confirm our claims about the higher efficiency of agents with memory. Moreover, an extensive comparison with reinforcement learning agents highlights the advantages of motivated learning over reinforcement learning.