Interpretability and Accuracy Analysis of Fuzzy Rule Based System Designed for Abalone
Prabhash Chandra1, Devendra Agarwal2, Praveen Kumar Shukla3
1Prabhash Chandra*, Department of Computer Science & Engineering, School of Engineering, Babu Banarasi Das University, Lucknow, India.
2Dr. Devendra Agarwal, School of Computer Science & Engineering, Babu Banarasi Das University, Lucknow, India.
3Dr. Praveen Kumar Shukla, Department of Information Technology, Babu Banarasi Das Northen India Institute of Technology, Lucknow, India.
Manuscript received on January 02, 2020. | Revised Manuscript received on January 15, 2020. | Manuscript published on January 30, 2020. | PP: 1335-1340 | Volume-8 Issue-5, January 2020. | Retrieval Number: E6131018520/2020©BEIESP | DOI: 10.35940/ijrte.E6131.018520
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Fuzzy Rule Based Systems are playing vital role in the implementation of human decision making. The development of interpretable Fuzzy Rule Based Systems with improved accuracy is a crucial research aspect in fuzzy based systems. Mamdani type fuzzy rule based systems are used to implement the proposed model. In this manuscript a FRBS is implemented with Guaje Open-Access Java based software. The interpretability and accuracy assessments are recorded on the different experiments with various rule generation methods, like Fuzzy decision tree and Wang Mendel method. The results are found satisfactory and a trade-off is handled between interpretability and accuracy. The major concern of the experimentation is number and type of fuzzy partitions. K-means and Hierarchical Fuzzy Partitions are used in the experiments with three and five number of fuzzy partitions.
Keywords: Fuzzy Ruled Based System (FRBS), Fuzzy Logic, Interpretability, Accuracy.
Scope of the Article: Autonomic Computing and Agent-Based Systems.