Abstract
In the research of rule extraction from neural networks,fidelity describes how well the rules mimic the behavior of a neural network whileaccuracy describes how well the rules can be generalized. This paper identifies thefidelity-accuracy dilemma. It argues to distinguishrule extraction using neural networks andrule extraction for neural networks according to their different goals, where fidelity and accuracy should be excluded from the rule quality evaluation framework, respectively.
Similar content being viewed by others
References
Andrews R, Diederich J, Tickle A B. Survey and critique of techniques for extracting rules from trained artificial neural networks.Knowledge-Based Systems, 1995, 8(6): 373–389.
Tickle A B, Andrews R, Golea Met al. The truth will come to light: Directions and challenges in extracting the knowledge embedded within trained artificial neural networks.IEEE Trans. Neural Networks, 1998, 9(6): 1057–1067.
Gallant S I. Connectionist expert systems.Communications of the ACM, 1988, 31(2): 152–169.
Fu L. Rule learning by searching on adapted nets. InProc. the 9th National Conference on Artificial Intelligence, Anaheim, CA, 1991, pp.590–595.
Thrun S. Extracting rules from artificial neural networks with distributed representations. InAdvances in Neural Information Processing Systems 7, Tesauro G, Touretzky D, Leen T (Eds.), Cambridge, MA, MIT Press, 1995, pp.505–512.
Craven M W, Shavlik J W. Extracting tree-structured representations of trained networks. InAdvances in Neural Information Processing Systems 8, Touretzky D, Mozer M, Hasselmo M (Eds.), Cambridge, MA, MIT Press, 1996, pp.24–30.
Krishnan R. A systematic method for decompositional rule extraction from neural networks. InProc. the NIPS'96 Workshop on Rule Extraction from Trained Artificial Neural Networks, Queensland, Australia, 1997, pp.38–45.
Setiono R. Extracting rules from neural networks by pruning and hidden-unit splitting.Neural Computation, 1997, 9(1): 205–225.
Zhou Z H, Chen S F, Chen Z Q. A statistics based approach for extracting priority rules from trained neural networks. InProc. the IEEE-INNS-ENNS International Joint Conference on Neural Networks, Como, Italy, 2000, 3: 401–406.
Masuoka R, Watanabe N, Kawamura Aet al. Neurofuzzy systems_— Fuzzy inference using a structured neural network. InProc. the International Conference on Fuzzy Logic and Neural Networks, Iizuka, Japan, 1990, pp.173–177.
Mitra S. Fuzzy MLP based expert system for medical diagnosis.Fuzzy Sets and Systems, 1994, 65(2–3): 285–296.
Castro J L, Mantas C J, Benitez J M. Interpretation of artificial neural networks by means of fuzzy rules.IEEE Trans. Neural Networks, 2002, 13(1): 101–116.
Giles C L, Miller C B, Chen Det al. Learning and extracting finite state automata with second-order recurrent neural networks.Neural Computation, 1992, 4(3): 393–405.
Omlin C W, Giles C L, Miller C B. Heuristics for the extraction of rules from discrete time recurrent neural networks. InProc. the International Joint Conference on Neural Networks, Baltimore, MD, 1992, Vol. 1, pp.33–38.
Giles C L, Omlin C W. Extraction, insertion, and refinement of symbolic rules in dynamically driven recurrent networks.Connection Science, 1993, 5(3–4): 307–328.
Saito K, Nakano R. Extracting regression rules from neural networks.Neural Networks, 2002, 15(10): 1279–1288.
Setiono R, Leow W K, Zurada J M. Extraction of rules from artificial neural networks for nonlinear regression.IEEE Trans. Neural Networks, 2002, 13(3): 564–577.
Tickle A B, Orlowski M, Diederich J. DEDEC: A methodology for extracting rule from trained artificial neural networks. InProc. the AISB'96 Workshop on Rule Extraction from Trained Neural Networks, Brighton, UK, 1996, pp.90–102.
Craven M W, Shavlik J W. Using sampling and queries to extract rules from trained neural networks. InProc. the 11th Int. Conf. Machine Learning, New Brunswick, NJ, 1994, pp.37–45.
Golea M. On the complexity of rule extraction from neural networks and network querying. InProc. the AISB'96 Workshop on Rule Extraction from Trained Neural Networks, Brighton, UK, 1996, pp.51–59.
Roy A. On connectionism, rule extraction, and brainlike learning.IEEE Trans. Fuzzy Systems, 2000, 8(2): 222–227.
Duch W, Adamczak R, Grabczewski K. A new methodology of extraction, optimization and application of crisp and fuzzy logical rules.IEEE Trans. Neural Networks, 2001, 12(2): 277–306.
Zhou Z H, Jiang Y, Chen S F. Extracting symbolic rules from trained neural network ensembles.AI Communications, 2003, 16(1): 3–15.
Towell G, Shavlik J. The extraction of refined rules from knowledge based neural networks.Machine Learning, 1993, 13(1): 71–101.
Fu L. Rule generation from neural networks.IEEE Trans. Systems, Man and Cybernetics, 1994, 24(8): 1114–1124.
Craven M W, Shavlik J W. Extracting comprehensible concept representations from trained neural networks. InWorking Notes on the IJCAI'95 Workshop on Comprehensibility in Machine Learning, Montreal, Canada, 1995, pp.61–75.
Taha I A, Ghosh J. Symbolic interpretation of artificial neural networks.IEEE Trans. Knowledge and Data Engineering, 1999, 11(3): 448–463.
Setiono R. Extracting M-of-N rules from trained neural networks.IEEE Trans. Neural Networks, 2000, 11(2): 512–519.
Wolpert D H, Macready W G. No free lunch theorems for optimization.IEEE Trans. Evolutionary Computation, 1997, 1(1): 67–82.
Quinlan J R. Comparing connectionist and symbolic learning methods. InComputational Learning Theory and Natural Learning Systems, Rivest R L (Ed.), Vol. 1, Cambridge, MA, MIT Press, 1994, pp.445–456.
Chalup S, Hayward R, Diederich J. Rule extraction from artificial neural networks trained on elementary number classification tasks. InProc. the 9th Australian Conference on Neural Networks, Brisbane, Australia, 1998, pp.265–270.
Maire F. Rule-extraction by backpropagation of polyhedra.Neural Networks, 1999, 12(4–5): 717–725.
Bologna G. Rule extraction from a multi layer perceptron with staircase activation functions. InProc. the IEEE-INNS-ENNS International Joint Conference on Neural Networks, Como, Italy, 2000, 3: 419–424.
Vahed A, Omlin C W. Rule extraction from recurrent neural networks using a symbolic machine learning algorithm. InProc. the 6th International Conference on Neural Information Processing, Dunedin, New Zealand, 1999, pp.712–717.
Author information
Authors and Affiliations
Corresponding author
Additional information
This work was supported by the National Outstanding Youth Foundation of China under Grant No.60325237 and the National Natural Science Foundation of China under Grant No.60273033.
Rights and permissions
About this article
Cite this article
Zhou, ZH. Rule extraction: Using neural networks or for neural networks?. J. Comput. Sci. & Technol. 19, 249–253 (2004). https://doi.org/10.1007/BF02944803
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02944803