Lange, Jordy de (2023) Can predatory pressure explain flocking behavior. Master's Thesis / Essay, Artificial Intelligence.
|
Text
1695706312889_Thesis.pdf Download (2MB) | Preview |
|
![[img]](https://fse.studenttheses.ub.rug.nl/style/images/fileicons/text.png) |
Text
toestemming.pdf Restricted to Registered users only Download (100kB) |
Abstract
Flocking-like behavior occurs in many species. Swarms of birds, schools of fish and herds of large herbivores all display behavior that can be categorized as flocking. Research has been done not only into how this flocking works, but also to find out why flocking is prevalent. One theory is that flocking-like behavior helps in evading predators by confusing them. Hahn et al. (2019) showed that in a 2D environment, simple learning agents that observe their neighbour agents learn to display flocking-like behavior when hunted by a predator. This predator is able to be confused by the presence of multiple agents, losing hunting efficiency. This thesis attempts to extend this principle into a 3D environment. First, a replication of the 2D model from Hahn et al. (2019) was constructed and tested. Next, a 3D version of the same model was built in to order see if this flocking-like behavior still emerges in a 3D environment. The replication of the earlier study was unsuccessful. The agents in our 2D experiments display less flocking-like behavior than expected based on the earlier research. Any behavior resembling flocking can be explained through the agents simply fleeing the predator, instead of them observing their neighbours and forming a flock. The 3D experiment features a similar conclusion, with all flocking-like behaviour that we observed being explainable by the agents
| Item Type: | Thesis (Master's Thesis / Essay) |
|---|---|
| Supervisor name: | Kosinka, J. and Verbrugge, L.C. |
| Degree programme: | Artificial Intelligence |
| Thesis type: | Master's Thesis / Essay |
| Language: | English |
| Date Deposited: | 27 Sep 2023 06:53 |
| Last Modified: | 27 Sep 2023 06:53 |
| URI: | https://fse.studenttheses.ub.rug.nl/id/eprint/31469 |
Actions (login required)
 |
View Item |