%0 Thesis
%9 Skripsi
%A LUTFIA LILIN KHARIROH, NIM. 12650039
%B FAKULTAS SAINS DAN TEKNOLOGI
%D 2016
%F digilib:22237
%I UIN SUNAN KALIJAGA YOGYAKARTA
%K Backpropagation, Prediction, Salon and Spa, Treatment
%T JARINGAN SYARAF TIRUAN BACKPROPAGATION UNTUK MEMPREDIKSI LAYANAN PERAWATAN DI MAIRA SALON DAN SPA
%U https://digilib.uin-suka.ac.id/id/eprint/22237/
%X Data and information on care services is very important for the beauty salon. The data is used for planning the future care services, for example: the amount of drug which will be used for treatment, the number of reliable therapist to perform care services, and many other case. Prediction of hair care service aims to find out how much in percentage the hair care services taken by visitors in a week.  The data used in this study is the data of hair care services (cream bath, hair mask, hair spa and hair rejuvenation) by summing all data services. At early stage, researcher normalizes data using minmax algorithm, and determine the most optimal backpropagation network is by varying the activation function in the hidden layer to obtain the greatest percentage of success. After obtaining the most optimal network conducted training on preliminary data. Then researcher performs test on the new data with the most optimal network.  Overall the data used for training amounted to 110. The Data we use is 77 data or 70% of the overall data. Testing data is 33 data represent 30% of overall data. Networks used in prediction using tansig-purelin activation function, with a hidden layer node layer 5, the epoch of 1000, learning rate 0.9 and momentum 0.7. The result is prediction appropriate 66.66% with tolerance value 3 and for unappropriated data presentation is 33.33%.
%Z Nurochman, M.Kom