Evaluation of Students Performance using Hierarchical Fuzzy Inference System

Abstract

Fuzzy Inference Systems (FIS) has often been used to evaluate performance using few input variables as a result of fear for rules explosion. This problem is solved using Hierarchical Fuzzy Inference System (HFIS); a divide-andconquer approach that drastically reduce the number of rules at the same time preserved the fuzzy logic reasoning. As a result, this study explore the potential of this tool in details by applying it to evaluate students' exam records. The proposed model is compared to classical one and results show that HFIS is more promising from the perspective of simplicity and precision. However, for optimum results, the study suggests training FIS with neural networks and emerging optimization algorithms.

Enhanced Approach for Change of Course of Study Using Fuzzy Logic

“…Fuzzy Inference System (FIS), constitute the expert' knowledge and experience towards designing a fuzzy logic system, which control the mappings of inputs and output on the set of define fuzzy rules (Yager, 1993). FIS has been a tool used to evaluate performance of a system with few inputs, in this problem due to computations of large inputs, the Hierarchical Fuzzy Inference System (HFIS) is employed to drastically reduced the large amount of rules expected in the system to a reasonable amount and at the same time preserved the accuracy behind the logic (Abdulkadir, et al, 2019). Fuzzy Inference System mainly consists of four stages namely: Fuzzification, Inference Engine, Rule Base and Defuzzication.…”