Reinforcement learning for ecosystems- Using explainable dynamic neural networks to train reinforcement learning agents in a simulated virtual animal ecosystem

Examensarbete för masterexamen

Please use this identifier to cite or link to this item: https://hdl.handle.net/20.500.12380/300934
Download file(s):
File Description SizeFormat 
Master Thesis Felix Hultén.pdf10.08 MBAdobe PDFView/Open
Type: Examensarbete för masterexamen
Title: Reinforcement learning for ecosystems- Using explainable dynamic neural networks to train reinforcement learning agents in a simulated virtual animal ecosystem
Authors: Hulthén, Felix
Abstract: This report covers the development of an inherently explainable and dynamic neural network for reinforcement learning in virtual animal ecosystems, based on the lifelong learning from zero (LL0) neural network used in supervised learning. The developed network (RLL0) is a fuzzy neural network with specialised growing and pruning rules for an ever changing environment. Results from benchmarking against a reference network show that RLL0 has a comparable performance while using far fewer trainable parameters. This, combined with its adapted architecture for visualising a learnt behaviour, shows promising future extensions to and use of the explored algorithms for animal behavioural based biological simulations.
Keywords: Life-long learning, reinforcement learning, fuzzy neural networks, artificial intelligence, animals
Issue Date: 2020
Publisher: Chalmers tekniska högskola / Institutionen för matematiska vetenskaper
URI: https://hdl.handle.net/20.500.12380/300934
Collection:Examensarbeten för masterexamen // Master Theses



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.