Hybrid expert system using case based reasoning and neural network for classification

Abstract Case Based Reasoning (CBR) is an analogical reasoning method, which solves problems by relating some previously solved problems to a current unsolved problem to draw analogical inferences for problem solving. But CBR faces the challenge of assigning weights to the features to measure similarity between a current unsolved case and cases stored in the case base effectively and correctly. The concept of neural network’s pruning is already used to sort out feature weighting problem in CBR. But it loses generality and actual elicited knowledge in the ANN’s links. This work proposes a method to extract symbolic weights from a trained neural network by observing the whole trained neural network as an AND/OR graph and then finds solution for each node that becomes the weight of a corresponding node. The proposed feature weighting mechanism is used in CBR to build a hybrid expert system for classification task and the performance of the proposed hybrid system is compared with that with other feature weighting mechanisms. The performance is validated on swine flu dataset and ionosphere, sonar and heart datasets collected from UCI repository. From the empirical results it is observed that in all the experiments the proposed feature weighting mechanism outperforms most of the earlier weighting mechanisms extracted from trained neural network.

[1]  Sang-Chan Park,et al.  A hybrid approach of neural network and memory-based learning to data mining , 2000, IEEE Trans. Neural Networks Learn. Syst..

[2]  Farhi Marir,et al.  Case-based reasoning: A review , 1994, The Knowledge Engineering Review.

[3]  Sang-Chan Park,et al.  Case-based reasoning and neural network based expert system for personalization , 2007, Expert Syst. Appl..

[4]  Sungho Ha,et al.  A personalized counseling system using case-based reasoning with neural symbolic feature weighting (CANSY) , 2008, Applied Intelligence.

[5]  Xining Li,et al.  An object-oriented And/Or graph inference engine , 1993, Proceedings of Canadian Conference on Electrical and Computer Engineering.

[6]  Sang-Chan Park,et al.  MBNR: Case-Based Reasoning with Local Feature Weighting by Neural Network , 2004, Applied Intelligence.

[7]  Huan Liu,et al.  Neural-network feature selector , 1997, IEEE Trans. Neural Networks.

[8]  Baohua Yang,et al.  Method on Determining Feature Weight in Case-Based Reasoning System , 2010, 2010 International Conference on E-Product E-Service and E-Entertainment.

[9]  Xinchuan Zeng,et al.  Feature weighting using neural networks , 2004, 2004 IEEE International Joint Conference on Neural Networks (IEEE Cat. No.04CH37541).

[10]  Yan Peng,et al.  A Case-based Reasoning with Feature Weights Derived by BP Network , 2007 .

[11]  Sang-Chan Park,et al.  Feature-Weighted CBR with Neural Network for Symbolic Features , 2006, ICIC.

[12]  Shreyash Tambe,et al.  EFFECTIVE DATA MINING USING NEURAL NETWORKS , 2016 .

[13]  Thomas G. Dietterich,et al.  An experimental comparison of the nearest-neighbor and nearest-hyperrectangle algorithms , 1995, Machine Learning.

[14]  Ms. Ishtake.  " Intelligent Heart Disease Prediction System Using Data Mining Techniques " , .

[15]  Miquel Sànchez-Marrè,et al.  A comparative study on the use of similarity measures in case-based reasoning to improve the classification of environmental system situations , 2004, Environ. Model. Softw..

[16]  Sung Ho Ha,et al.  A personalized counseling system using case-based reasoning with neural symbolic feature weighting (CANSY) , 2007, Applied Intelligence.

[17]  Sang C. Park,et al.  Memory- and neural-network-based expert system , 1999, Defense, Security, and Sensing.

[18]  Zia Ul-Qayyum,et al.  CBR and Neural Networks Based Technique for Predictive Prefetching , 2010, MICAI.

[19]  Jude W. Shavlik,et al.  Using neural networks for data mining , 1997, Future Gener. Comput. Syst..

[20]  David Leake,et al.  Case-Based Reasoning: Experiences, Lessons and Future Directions , 1996 .

[21]  Lei Wang,et al.  Knowledge Points Organization Model based on AND/OR Graph in ICAI , 2010, 2010 Sixth International Conference on Natural Computation.

[22]  Jacques de Villiers,et al.  Backpropagation neural nets with one and two hidden layers , 1993, IEEE Trans. Neural Networks.

[23]  Agnar Aamodt,et al.  Case-Based Reasoning: Foundational Issues, Methodological Variations, and System Approaches , 1994, AI Commun..

[24]  David W. Aha,et al.  A Review and Empirical Evaluation of Feature Weighting Methods for a Class of Lazy Learning Algorithms , 1997, Artificial Intelligence Review.

[25]  Walter Daelemans,et al.  The Acquisition of Stress: A Data-Oriented Approach , 1994, Comput. Linguistics.

[26]  Yan Peng,et al.  A Case-based Reasoning with Feature Weights Derived by BP Network , 2007, Workshop on Intelligent Information Technology Application (IITA 2007).