Medical diagnostic expert system based on PDP model

Saito,; Nakano, Mikio

doi:10.1109/icnn.1988.23855

Cited by 148 publications

(73 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…To limit the number of nodes in the search graph one may try to limit the number of literals in the antecedents of extracted rules. In one of the first neural rule extraction methods Saito and Nakano [14] restricted the maximum number of positive and negative literals and the depth of the breadth-first search process, additionally restricting the search tree to those combinations of literals that were present in the training set. Due to these restrictions their method could sometimes accept a rule that was too general.…”

Section: An Overview Of Neural Rule Extraction Methodsmentioning

confidence: 99%

“…For classification tasks, when crisp logical decisions are required, the prior probability of the weight values should include not only small weights, but also large positive and negative weights distributed around 1. For example (14) where the parameters play a similar role for probabilities as the parameters for the cost function. Using alternative cost functions amounts to different priors for regularization, for example using Laplace instead of the Gaussian prior.…”

Section: A Mlp2ln: Changing Mlp Into Logical Networkmentioning

confidence: 99%

See 1 more Smart Citation

A new methodology of extraction, optimization and application of crisp and fuzzy logical rules

Duch

Adamczak

Grąbczewski

2001

IEEE Trans. Neural Netw.

204

155

View full text Add to dashboard Cite

Abstract-A new methodology of extraction, optimization, and application of sets of logical rules is described. Neural networks are used for initial rule extraction, local, or global minimization procedures for optimization, and Gaussian uncertainties of measurements are assumed during application of logical rules. Algorithms for extraction of logical rules from data with real-valued features require determination of linguistic variables or membership functions. Context-dependent membership functions for crisp and fuzzy linguistic variables are introduced and methods of their determination described. Several neural and machine learning methods of logical rule extraction generating initial rules are described, based on constrained multilayer perceptron, networks with localized transfer functions or on separability criteria for determination of linguistic variables. A tradeoff between accuracy/simplicity is explored at the rule extraction stage and between rejection/error level at the optimization stage. Gaussian uncertainties of measurements are assumed during application of crisp logical rules, leading to "soft trapezoidal" membership functions and allowing to optimize the linguistic variables using gradient procedures. Numerous applications of this methodology to benchmark and real-life problems are reported and very simple crisp logical rules for many datasets provided.

show abstract

Section: An Overview Of Neural Rule Extraction Methodsmentioning

confidence: 99%

Section: A Mlp2ln: Changing Mlp Into Logical Networkmentioning

confidence: 99%

A new methodology of extraction, optimization and application of crisp and fuzzy logical rules

Duch

Adamczak

Grąbczewski

2001

IEEE Trans. Neural Netw.

204

155

View full text Add to dashboard Cite

show abstract

“…A large number of medical diagnosis systems is also available, including DiagnosMD, DXPlain (London, 1998), eMedicine, Isabel (Ramnarayan et al, 2003), IWSMD (Ibrahim, AbdelRahman, & Farag, 2008), EasyDiagnosis, ADM (Le Beux et al, 1991), and Your Diagnosis. Other types of sys tems exist, such as those based on the PDP model (Saito & Nakano, 1988).…”

Section: Related Workmentioning

confidence: 99%

ODDIN: Ontology-driven differential diagnosis based on logical inference and probabilistic refinements

García-Crespo

Rodríguez

Mencke

et al. 2010

Expert Systems with Applications

View full text Add to dashboard Cite

a b s t r a c tMedical differential diagnosis (ddx) is based on the estimation of multiple distinct parameters in order to determine the most probable diagnosis. Building an intelligent medical differential diagnosis system implies using a number of knowledge based technologies which avoid ambiguity, such as ontologies rep resenting specific structured information, but also strategies such as computation of probabilities of var ious factors and logical inference, whose combination outperforms similar approaches. This paper presents ODDIN, an ontology driven medical diagnosis system which applies the aforementioned strat egies. The architecture and proof of concept implementation is described, and results of the evaluation are discussed.

show abstract

“…The method we introduce for extracting rules from neural networks, as well as those previously described (e.g., Saito & Nakano, 1988;Fu, 1991), does so by trying to find combinations of the input values to a unit that result in it having an activation near one. Hence, in this section we describe the mathematics underlying the neural networks that we use.…”

Section: Commonalitiesmentioning

confidence: 99%

Untitled

1993

View full text Add to dashboard Cite

Abstract. Neural networks, despite their empirically proven abilities, have been little used for the refinement of existing knowledge because this task requires a three-step process. First, knowledge must be inserted into a neural network. Second, the network must be refined. Third, the refined knowledge must be extracted from the network. We have previously described a method for the first step of this process. Standard neural learning techniques can accomplish the second step. In this article, we propose and empirically evaluate a method for the final, and possibly most difficult, step. Our method efficiently extracts symbolic rules from trained neural networks. The four major results of empirical tests of this method are that the extracted rules 1) closely reproduce the accuracy of the network from which they are extracted; 2) are superior to the rules produced by methods that directly refine symbolic rules; 3) are superior to those produced by previous techniques for extracting rules from trained neural networks; and 4) are "human comprehensible." Thus, this method demonstrates that neural networks can be used to effectively refine symbolic knowledge. Moreover, the rule-extraction technique developed herein contributes to the understanding of how symbolic and connectionist approaches to artificial intelligence can be profitably integrated.

show abstract

Medical diagnostic expert system based on PDP model

Cited by 148 publications

References 5 publications

A new methodology of extraction, optimization and application of crisp and fuzzy logical rules

A new methodology of extraction, optimization and application of crisp and fuzzy logical rules

ODDIN: Ontology-driven differential diagnosis based on logical inference and probabilistic refinements

Untitled

Contact Info

Product

Resources

About