A Combined Multi-Objective Memetic Algorithm and ANFIS for Heat Stroke Prediction

Abstract

Heat stroke risk prediction is a problem that
demands high classification accuracy. It is a tool to help
prevent heat stroke occurrence. This research introduced a
prediction system based on a combination of Multi-
Objective Memetic Algorithm (MOMA) and Adaptive
Neuro-Fuzzy Inference System (ANFIS) in brief MOMAANFIS.
When MOMA-ANFIS was applied to solve the
problem in the prediction of heat stroke risks, it was
found that the accuracy rate of the classification test
result was as high as 98.51% which was greater than the
rate obtained when the traditional ANFIS. As the number
of rules decreased, the fuzzy rule architecture became less
complicated.