Kombinasi Multilayer Fuzzy Inference System (MFIS) dengan K-Means pada Sistem Klasifikasi Penyakit Gigi

Abstract

Screening is an important procedure that must be passed in every task before determining the action to be taken by all health workers. In the field of dental and oral health, a patient is often found who cannot represent his condition. Uncertainty about the explanation of the level of pain conveyed by a patient often hinders the examination process and the actions to be taken. To resolve the uncertainty of the information then built a system by applying fuzzy rules. Based on the analysis of system requirements, there are several multilevel questions that can stop at the initial stage without having to be fulfilled by other indicators. To answer this problem, a fuzzy system is built using the multilayer concept where if the needs of the first layer have been met then the system will not continue to the second layer, conversely if results are still not found in the first layer, then the system will continue to the second layer to perform algorithmic calculations. to complete the task. The classification system certainly has a weakness, namely it can only solve problems that have previously been taught on training data. In this study, problems with the classification system were solved using the K-Means method. The clustering concept is used in the second layer to get even better results. Based on the results of research that has been done, the combination of the multilayer fuzzy inference system (MFIS) with the K-Means method in the dental disease classification system can work well and carry out its functions with optimal results.