Induction of Finite-State Automata Using Second-Order Recurrent Networks

By Raymond L. Watrous, Gary M. Kuhn. 1991

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Deep learning in neural networks: An overview (2014, methodology): “LSTM clearly outperformed previous RNNs on tasks that require learning the rules of regular languages describable by deterministic Finite State Automata (FSAs) (Watrous and Kuhn, 1992; Casey, 1996; Siegelmann, 1992; Blair and Pollack, 1997; Kalinke and Lehmann, 1998; Zeng et al., 1994; Manolios and ”
Recurrent neural networks for time series classification (2003, background): “Recurrent neural networks (RNNs) are dynamical systems that make efficient use of temporal information in the input sequence, both for classification [6,9,17] as well as for prediction [3,13].”
Learning Finite State Representations of Recurrent Policy Networks (2018, unknown): “& Kuhn, 1992; Weiss et al., 2017)).”
First-Order Recurrent Neural Networks and Deterministic Finite State Automata (1994, result): “These results are comparable to those obtained with secondorder networks (Giles et al. 1992; Watrous and Kuhn 1992a; Zeng et al. 1993)., Watrous and Kuhn (1992a) suggest that the reason why some of their networks are not able to generalize is that the grammar is not sufficiently constrained by th”
Rule Extraction from Recurrent Neural Networks: ATaxonomy and Review (2005, methodology): “…in the RNN (see Section 4 for more complete reference lists): searching (Giles et al. 1991, Zeng et al. 1993, Frasconi et al. 1996) and sampling (Watrous & Kuhn 1992, Manolios & Fanelli 1994, Alqu´ezar & Sanfeliu 1994 a , Tiˇno & ˇSajda 1995, Schellhammer et al. 1998, Tiˇno & Vojtek 1998, Tiˇno…”
Bridging Long Time Lags by Weight Guessing and \Long Short Term Memory" (1996, background): “Some of Tomita's grammars [30] are also often used as benchmark problems for recurrent nets [2, 31, 23, 17].”
An Empirical Evaluation of Rule Extraction from Recurrent Neural Networks (2017, background): “Also, prior work (Giles et al., 1992; Watrous & Kuhn, 1992; Omlin & Giles, 1996c; Casey, 1996; Jacobsson, 2005) has shown that it is possible to extract deterministic ﬁnite automata (DFA) from RNN models trained to perform grammatical inference and that grammatical rules can be stably encoded in…”