PERBANDINGAN TIPE FUNGSI KEANGGOTAAN ANFIS UNTUK MENGUKUR AKURASI EVALUASI KINERJA DOSEN

Abstract

Lecturer is one essential component in a system of education in universities. To improve teaching quality one of which can be done is to conduct an evaluation of the performance of lecturers. Teaching evaluation affects the learning process in educational environment, errors in evaluating the performance of one of them is causing the decline in teacher motivation in the teaching faculty. This will affect the learning process as a whole.In this research approach Adaptive Neuro Fuzzy Inference System is used to process and improve accuracy IPKAD faculty performance assessment. Several types of membership functions tested to compare the accuracy of the resulting assessment. Tests using a different type of membership function produces different accuracy. In this study indicate that this type of triangular membership functions and trapezium produces a higher level of accuracy compared with Gaussian membership function type and bells.