39
IRUS TotalDownloads
Altmetric
Verification of RNN-based neural agent-environment systems
| File | Description | Size | Format | |
|---|---|---|---|---|
 AAAI-AkintundeM.pdf | Accepted version | 241.71 kB | Adobe PDF | View/Open |
| Title: | Verification of RNN-based neural agent-environment systems |
| Authors: | Akintunde, M Kevorchian, A Lomuscio, A Pirovano, E |
| Item Type: | Conference Paper |
| Abstract: | We introduce agent-environment systems where the agent is stateful and executing a ReLU recurrent neural network. We define and study their verification problem by providing equivalences of recurrent and feed-forward neural networks on bounded execution traces. We give a sound and complete procedure for their verification against properties specified in a simplified version of LTL on bounded executions. We present an implementation and discuss the experimental results obtained. |
| Issue Date: | 17-Jul-2019 |
| Date of Acceptance: | 1-Nov-2018 |
| URI: | http://hdl.handle.net/10044/1/77521 |
| ISSN: | 2159-5399 |
| Publisher: | Association for the Advancement of Artificial Intelligence |
| Journal / Book Title: | Proceedings of the ... AAAI Conference on Artificial Intelligence. AAAI Conference on Artificial Intelligence |
| Volume: | 33 |
| Issue: | 1 |
| Copyright Statement: | Copyright © 2019 Association for the Advancement of Artificial Intelligence |
| Sponsor/Funder: | Royal Academy Of Engineering Defence Advanced Research Projects Agency (UK) Royal Academy Of Engineering |
| Funder's Grant Number: | CIET 1718/26 Ref: FA8750-18-C-0095 CIET1819\TUA\20 |
| Conference Name: | 33rd AAAI Conference on Artificial Intelligence |
| Publication Status: | Published |
| Start Date: | 2019-01-27 |
| Finish Date: | 2019-02-01 |
| Conference Place: | Honolulu,Hawaii, USA |
| Appears in Collections: | Computing |