| dc.date.accessioned |
2014-06-03T11:30:35Z |
und |
| dc.date.accessioned |
2017-10-24T12:23:48Z |
|
| dc.date.available |
2014-06-03T11:30:35Z |
und |
| dc.date.available |
2017-10-24T12:23:48Z |
|
| dc.date.issued |
2014-06-03T11:30:35Z |
|
| dc.identifier.uri |
http://radr.hulib.helsinki.fi/handle/10138.1/3756 |
und |
| dc.identifier.uri |
http://hdl.handle.net/10138.1/3756 |
|
| dc.title |
Evolving Agents using NEAT to Achieve Human-Like Play in FPS Games |
en |
| ethesis.discipline |
Computer science |
en |
| ethesis.discipline |
Tietojenkäsittelytiede |
fi |
| ethesis.discipline |
Datavetenskap |
sv |
| ethesis.discipline.URI |
http://data.hulib.helsinki.fi/id/1dcabbeb-f422-4eec-aaff-bb11d7501348 |
|
| ethesis.department.URI |
http://data.hulib.helsinki.fi/id/225405e8-3362-4197-a7fd-6e7b79e52d14 |
|
| ethesis.department |
Institutionen för datavetenskap |
sv |
| ethesis.department |
Department of Computer Science |
en |
| ethesis.department |
Tietojenkäsittelytieteen laitos |
fi |
| ethesis.faculty |
Matematisk-naturvetenskapliga fakulteten |
sv |
| ethesis.faculty |
Matemaattis-luonnontieteellinen tiedekunta |
fi |
| ethesis.faculty |
Faculty of Science |
en |
| ethesis.faculty.URI |
http://data.hulib.helsinki.fi/id/8d59209f-6614-4edd-9744-1ebdaf1d13ca |
|
| ethesis.university.URI |
http://data.hulib.helsinki.fi/id/50ae46d8-7ba9-4821-877c-c994c78b0d97 |
|
| ethesis.university |
Helsingfors universitet |
sv |
| ethesis.university |
University of Helsinki |
en |
| ethesis.university |
Helsingin yliopisto |
fi |
| dct.creator |
Laster, Zachary Howell |
|
| dct.issued |
2014 |
|
| dct.language.ISO639-2 |
eng |
|
| dct.abstract |
Artificial agents are commonly used in games to simulate human opponents. This allows players to enjoy games without requiring them to play online or with other players locally. Basic approaches tend to suffer from being unable to adapt strategies and often perform tasks in ways very few human players could ever achieve. This detracts from the immersion or realism of the gameplay. In order to achieve more human-like play more advanced approaches are employed in order to either adapt to the player's ability level or to cause the agent to play more like a human player can or would.
Utilizing artificial neural networks evolved using the NEAT methodology, we attempt to produce agents to play a FPS-style game. The goal is to see if the approach produces well-playing agents with potentially human-like behaviors. We provide a large number of sensors and motors to the neural networks of a small population learning through co-evolution.
Ultimately we find that the approach has limitations and is generally too slow for practical application, but holds promise for future developments. Many extensions are presented which could improve the results and reduce training times. The agents learned to perform some basic tasks at a very rough level of skill, but were not competitive at even a beginner level. |
en |
| dct.language |
en |
|
| ethesis.language.URI |
http://data.hulib.helsinki.fi/id/languages/eng |
|
| ethesis.language |
English |
en |
| ethesis.language |
englanti |
fi |
| ethesis.language |
engelska |
sv |
| ethesis.thesistype |
pro gradu-avhandlingar |
sv |
| ethesis.thesistype |
pro gradu -tutkielmat |
fi |
| ethesis.thesistype |
master's thesis |
en |
| ethesis.thesistype.URI |
http://data.hulib.helsinki.fi/id/thesistypes/mastersthesis |
|
| dct.identifier.urn |
URN:NBN:fi-fe2017112252510 |
|
| dc.type.dcmitype |
Text |
|
