Adaptive Behavior in Artificial Intelligence
Camacho, Claudio (2009)
Metropolia Ammattikorkeakoulu
2009
All rights reserved
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-200912047036
https://urn.fi/URN:NBN:fi:amk-200912047036
Tiivistelmä
The goal of this project was to study the characteristics and consequences of using dynamic behavior in artificial intelligence (AI). The main purpose was to create a
computer chess program with two different artificial intelligence engines and using them as a tool for performing different experiments.
The project was carried out by developing a simple chess
program with two different AI engines, one containing classical AI techniques without the use of dynamism, and
another AI engine with adaptive behavior and dynamic evaluation techniques.
The results showed a remarkable improvement on the adaptive AI engine over the classic AI engine. The results also demonstrated the challenge of implementing proper adaptive behavior for the system. Nevertheless, it was proved that it was feasible to program certain dynamism, as an extension to the classic AI engine, thus enabling the intelligent agent to provide adaptive and human-like behavior when playing a game.
To summarize, the results proved the possibility to create a simplified model of a computational intelligent system with certain basic dynamic behavior capabilities. Nonetheless, it is recommended to further develop the current implementation in order to enable the intelligent agent to widen its dynamic behavior functionality and to implement basic learning capabilities.
computer chess program with two different artificial intelligence engines and using them as a tool for performing different experiments.
The project was carried out by developing a simple chess
program with two different AI engines, one containing classical AI techniques without the use of dynamism, and
another AI engine with adaptive behavior and dynamic evaluation techniques.
The results showed a remarkable improvement on the adaptive AI engine over the classic AI engine. The results also demonstrated the challenge of implementing proper adaptive behavior for the system. Nevertheless, it was proved that it was feasible to program certain dynamism, as an extension to the classic AI engine, thus enabling the intelligent agent to provide adaptive and human-like behavior when playing a game.
To summarize, the results proved the possibility to create a simplified model of a computational intelligent system with certain basic dynamic behavior capabilities. Nonetheless, it is recommended to further develop the current implementation in order to enable the intelligent agent to widen its dynamic behavior functionality and to implement basic learning capabilities.