MODEL HYBRID JARINGAN SYARAF BACKPROPAGATION DAN REGRESI FUZZY UNTUK PERAMALAN DATA TIME SERIES

• Main Authors: , Winna leon Agusta, , Prof. Drs. H.Subanar, Ph.D
• 格式: Theses and Dissertations NonPeerReviewed
• 出版: [Yogyakarta] : Universitas Gadjah Mada 2013
• 主題: ETD
• 在線閱讀: https://repository.ugm.ac.id/118811/; http://etd.ugm.ac.id/index.php?mod=penelitian_detail&sub=PenelitianDetail&act=view&typ=html&buku_id=58787

In this thesis hybrid models Backpropagation of neural network and fuzzy regression model are formed to improve the effectiveness of the performance of fuzzy regression models using Backpropagation of neural network performance for forecasting the case of incomplete data or the use of data in very small amounts and short period. In the implementation procedure, the output of the optimization process Backpropagation of neural network used as input to the fuzzy regression to obtain the minimum fuzzy interval. In this study, empirical data analysis of monthly Consumer Price Index (CPI) data and exchange rates Rupiah used in the application of hybrid models are presented. Proposed hybrid model is able to provide the smallest possible interval that includes all the actual data in it. Thus presenting the hybrid model is intended to provide a reference for decision-makers to look at the best and worst possible conditions of the observations made