Matches in SemOpenAlex for { <https://semopenalex.org/work/W2000257498> ?p ?o ?g. }
- W2000257498 endingPage "93" @default.
- W2000257498 startingPage "78" @default.
- W2000257498 abstract "An important drawback of many artificial neural networks (ANN) is their lack of explanation capability [Andrews, R., Diederich, J., & Tickle, A. B. (1996). A survey and critique of techniques for extracting rules from trained artificial neural networks. Knowledge-Based Systems, 8, 373–389]. This paper starts with a survey of algorithms which attempt to explain the ANN output. We then present HYPINV,1 a new explanation algorithm which relies on network inversion; i.e. calculating the ANN input which produces a desired output. HYPINV is a pedagogical algorithm, that extracts rules, in the form of hyperplanes. It is able to generate rules with arbitrarily desired fidelity, maintaining a fidelity–complexity tradeoff. To our knowledge, HYPINV is the only pedagogical rule extraction method, which extracts hyperplane rules from continuous or binary attribute neural networks. Different network inversion techniques, involving gradient descent as well as an evolutionary algorithm, are presented. An information theoretic treatment of rule extraction is presented. HYPINV is applied to example synthetic problems, to a real aerospace problem, and compared with similar algorithms using benchmark problems." @default.
- W2000257498 created "2016-06-24" @default.
- W2000257498 creator A5011227186 @default.
- W2000257498 creator A5038037619 @default.
- W2000257498 date "2007-01-01" @default.
- W2000257498 modified "2023-09-26" @default.
- W2000257498 title "Neural network explanation using inversion" @default.
- W2000257498 cites W1511089533 @default.
- W2000257498 cites W1530478301 @default.
- W2000257498 cites W1539435948 @default.
- W2000257498 cites W1546185221 @default.
- W2000257498 cites W1547466200 @default.
- W2000257498 cites W1614345895 @default.
- W2000257498 cites W1928044808 @default.
- W2000257498 cites W1934057878 @default.
- W2000257498 cites W1968680127 @default.
- W2000257498 cites W1981249599 @default.
- W2000257498 cites W2005695783 @default.
- W2000257498 cites W2012612681 @default.
- W2000257498 cites W2023577742 @default.
- W2000257498 cites W2053987251 @default.
- W2000257498 cites W2063046703 @default.
- W2000257498 cites W2065204741 @default.
- W2000257498 cites W2069543264 @default.
- W2000257498 cites W2072199831 @default.
- W2000257498 cites W2073766055 @default.
- W2000257498 cites W2079397195 @default.
- W2000257498 cites W2082688942 @default.
- W2000257498 cites W2094438588 @default.
- W2000257498 cites W2094558429 @default.
- W2000257498 cites W2101743164 @default.
- W2000257498 cites W2103518983 @default.
- W2000257498 cites W2112036221 @default.
- W2000257498 cites W2118259313 @default.
- W2000257498 cites W2122925692 @default.
- W2000257498 cites W2136734237 @default.
- W2000257498 cites W2143945210 @default.
- W2000257498 cites W2144185668 @default.
- W2000257498 cites W2144212877 @default.
- W2000257498 cites W2144914634 @default.
- W2000257498 cites W2145470394 @default.
- W2000257498 cites W2150355110 @default.
- W2000257498 cites W2151523532 @default.
- W2000257498 cites W2151968358 @default.
- W2000257498 cites W2164551755 @default.
- W2000257498 cites W2172463260 @default.
- W2000257498 cites W2304602072 @default.
- W2000257498 cites W4235570552 @default.
- W2000257498 cites W4236137412 @default.
- W2000257498 cites W4244296837 @default.
- W2000257498 cites W2131974528 @default.
- W2000257498 doi "https://doi.org/10.1016/j.neunet.2006.07.005" @default.
- W2000257498 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/17029713" @default.
- W2000257498 hasPublicationYear "2007" @default.
- W2000257498 type Work @default.
- W2000257498 sameAs 2000257498 @default.
- W2000257498 citedByCount "77" @default.
- W2000257498 countsByYear W20002574982012 @default.
- W2000257498 countsByYear W20002574982013 @default.
- W2000257498 countsByYear W20002574982014 @default.
- W2000257498 countsByYear W20002574982015 @default.
- W2000257498 countsByYear W20002574982016 @default.
- W2000257498 countsByYear W20002574982017 @default.
- W2000257498 countsByYear W20002574982018 @default.
- W2000257498 countsByYear W20002574982019 @default.
- W2000257498 countsByYear W20002574982020 @default.
- W2000257498 countsByYear W20002574982021 @default.
- W2000257498 countsByYear W20002574982022 @default.
- W2000257498 countsByYear W20002574982023 @default.
- W2000257498 crossrefType "journal-article" @default.
- W2000257498 hasAuthorship W2000257498A5011227186 @default.
- W2000257498 hasAuthorship W2000257498A5038037619 @default.
- W2000257498 hasConcept C109007969 @default.
- W2000257498 hasConcept C11413529 @default.
- W2000257498 hasConcept C119857082 @default.
- W2000257498 hasConcept C13280743 @default.
- W2000257498 hasConcept C151730666 @default.
- W2000257498 hasConcept C153258448 @default.
- W2000257498 hasConcept C154945302 @default.
- W2000257498 hasConcept C185798385 @default.
- W2000257498 hasConcept C1893757 @default.
- W2000257498 hasConcept C205649164 @default.
- W2000257498 hasConcept C2524010 @default.
- W2000257498 hasConcept C2776459999 @default.
- W2000257498 hasConcept C33923547 @default.
- W2000257498 hasConcept C41008148 @default.
- W2000257498 hasConcept C48372109 @default.
- W2000257498 hasConcept C50644808 @default.
- W2000257498 hasConcept C68693459 @default.
- W2000257498 hasConcept C76155785 @default.
- W2000257498 hasConcept C86803240 @default.
- W2000257498 hasConcept C94375191 @default.
- W2000257498 hasConceptScore W2000257498C109007969 @default.
- W2000257498 hasConceptScore W2000257498C11413529 @default.
- W2000257498 hasConceptScore W2000257498C119857082 @default.
- W2000257498 hasConceptScore W2000257498C13280743 @default.
- W2000257498 hasConceptScore W2000257498C151730666 @default.
- W2000257498 hasConceptScore W2000257498C153258448 @default.