Constrained Training of Recurrent Neural Networks for Automata Learning

被引：1

作者：

Aichernig, Bernhard K. ^{[1
]}

Koenig, Sandra ^{[2
]}

Mateis, Cristinel ^{[2
]}

Pferscher, Andrea ^{[1
]}

Schmidt, Dominik ^{[2
]}

Tappler, Martin ^{[1
,3
]}

机构：

[1] Graz Univ Technol, Inst Software Technol, Graz, Austria

[2] AIT Austrian Inst Technol, Vienna, Austria

[3] Graz Univ Technol, Silicon Austria Labs, SAL DES Lab, Graz, Austria

来源：

SOFTWARE ENGINEERING AND FORMAL METHODS, SEFM 2022 | 2022年 / 13550卷

关键词：

Automata learning; Machine learning; Recurrent neural networks; Bluetooth Low Energy; Model inference;

D O I：

10.1007/978-3-031-17108-6_10

中图分类号：

TP31 [计算机软件];

学科分类号：

081202 ; 0835 ;

摘要：

In this paper, we present a novel approach to learning finite automata with the help of recurrent neural networks. Our goal is not only to train a neural network that predicts the observable behavior of an automaton but also to learn its structure, including the set of states and transitions. In contrast to previous work, we constrain the training with a specific regularization term. We evaluate our approach with standard examples from the automata learning literature, but also include a case study of learning the finite-state models of real Bluetooth Low Energy protocol implementations. The results show that we can find an appropriate architecture to learn the correct automata in all considered cases.

引用

页码：155 / 172

页数：18

共 50 条

[1] Learning minimal automata with recurrent neural networks
Aichernig, Bernhard K.
Koenig, Sandra
Mateis, Cristinel
Pferscher, Andrea
Tappler, Martin
SOFTWARE AND SYSTEMS MODELING, 2024, 23 (03): : 625 - 655
[2] Evaluating the Learning of Automata through the Use of Recurrent Neural Networks
Lima, L.
Sampaio, A.
IEEE LATIN AMERICA TRANSACTIONS, 2018, 16 (10) : 2609 - 2616
[3] Training and extraction of fuzzy finite state automata in recurrent neural networks
Chandra, Rohitash
Omlin, Christian W.
PROCEEDINGS OF THE SECOND IASTED INTERNATIONAL CONFERENCE ON COMPUTATIONAL INTELLIGENCE, 2006, : 271 - 275
[4] Training Deep Neural Networks with Constrained Learning Parameters
Date, Prasanna
Carothers, Christopher D.
Mitchell, John E.
Hendler, James A.
Magdon-Ismail, Malik
2020 INTERNATIONAL CONFERENCE ON REBOOTING COMPUTING (ICRC 2020), 2020, : 107 - 115
[5] Recurrent neural networks and finite automata
Siegelmann, HT
COMPUTATIONAL INTELLIGENCE, 1996, 12 (04) : 567 - 574
[6] Connecting Weighted Automata and Recurrent Neural Networks through Spectral Learning
Rabusseau, Guillaume
Li, Tianyu
Precup, Doina
22ND INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND STATISTICS, VOL 89, 2019, 89
[7] AdaAX: Explaining Recurrent Neural Networks by Learning Automata with Adaptive States
Hong, Dat
Segre, Alberto Maria
Wang, Tong
PROCEEDINGS OF THE 28TH ACM SIGKDD CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING, KDD 2022, 2022, : 574 - 584
[8] Local Structure Helps Learning Optimized Automata in Recurrent Neural Networks
Binas, Jonathan
Indiveri, Giacomo
Pfeiffer, Michael
2015 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2015,
[9] Toward Training Recurrent Neural Networks for Lifelong Learning
Sodhani, Shagun
Chandar, Sarath
Bengio, Yoshua
NEURAL COMPUTATION, 2020, 32 (01) : 1 - 35
[10] Connecting weighted automata, tensor networks and recurrent neural networks through spectral learning
Tianyu Li
Doina Precup
Guillaume Rabusseau
Machine Learning, 2024, 113 : 2619 - 2653

← 1 2 3 4 5 →