Rule Extraction from Recurrent Neural Networks: ATaxonomy and Review

By Henrik Jacobsson. 2005

182 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Control and Systems Engineering

How researchers see this paper

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

In Artificial Intelligence

In Control and Systems Engineering

Citations in the corpus

Neural-Symbolic Computing: An Effective Methodology for Principled Integration of Machine Learning and Reasoning (2019, background): “Most of them are discussed in the surveys [1, 24, 55].”
Explaining machine learning models in sales predictions (2017, methodology): “A review of explanation techniques for neural models is given in (Jacobsson, 2005).”
Explainable Artificial Intelligence: a Systematic Review (2020, background): “This category contains literature and systematic reviews devoted to specific classes of solutions for explainability, such as systems generating textual explanations [23], or constrained to specific AI techniques as, for instance, neural networks [24] (summary in table .”
Fundamentals of natural computing: an overview (2007, background): “Among open problems there is knowledge extraction [143,149], the modeling of higher human cognitive abilities, interaction between emotional and conceptual, and the unification of language and cognition in evolving and integrated systems [227].”
Extracting Automata from Recurrent Neural Networks Using Queries and Counterexamples (2017, methodology): “Indeed, several lines of work attempt to extract clear rules for their decisions (Jacobsson, 2005; Omlin & Giles, 1996; Cechin et al., 2003).”
Dimensions of Neural-symbolic Integration - A Structured Survey (2005, background): “[Jacobsson, 2005] gives a recent overview over extraction methods.”
Learning Finite State Representations of Recurrent Policy Networks (2018, background): “There have been efforts made in the past to understand the internals of Recurrent Networks (Karpa-thy et al., 2015; Arras et al., 2017; Strobelt et al., 2016; Murdoch & Szlam, 2017; Jacobsson, 2005; Omlin & Giles, 1996).”
An Empirical Evaluation of Rule Extraction from Recurrent Neural Networks (2017, methodology): “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…”
Electricity price forecasting: A review of the state-of-the-art with a look into the future (2014, methodology): “See Janczura and Weron (2012) for an efficient modification of the algorithm to overcome this problem (Matlab code is available from http://ideas.”
Extended Taxonomy of Rule Extraction Techniques and Assessment of KDRuleEx (2012, methodology): “Possible application of rule extraction is suggested by Andrews et al [6], Tickel et al [17] and Jacobsson [27] in the context of ANN and may be summarized as: 1.”
Explaining Classifications For Individual Instances (2008, methodology): “For a more complete review of knowledge extraction methods from ANN we refer the reader to [17], where taxonomy and criteria for evaluation of rule extraction methods from ANN are introduced, and [18], which concentrates mostly on finite-state machines extracted from recurrent ANN but covers also ot”