PENENTUAN MODEL RETURN HARGA SAHAM DENGAN MULTI LAYER FEED FORWARD NEURAL NETWORK MENGGUNAKAN ALGORITMA RESILENT BACKPROPAGATION (Studi Kasus : Harga Penutupan Saham Unilever Indonesia Tbk. Periode September 2007 – Maret 2015)

Riza Adi Priantoro, Dwi Ispriyanti, Moch. Abdul Mukid

Abstract


Determination of a return of stock price model is often associated with a process of forecasting for future periods.  A method that can be used is neural network. The use of neural network in the field of forecasting can be a good solution, but the problem is how to determine the network architecture and the selection of appropriate training methods. One possible option is to use resilent back propagation algorithm. Resilent back propagation algorithm is a supervised learning algorithm to change the weights of the layers. This algorithm uses the error in the backward direction (back propagation), but previously performed advanced stage (feed forward) to get the error. This algorithm can be used as a learning method in training model of a multi-layer feed forward neural network. From the results of the training and testing on the share return of stock price PT. Unilever Indonesia Tbk. data obtained MSE value of 0.0329. This model is good to use because it provides a fairly accurate prediction of the results shown by the proximity of the target with the output.

Keywords : return, neural network, back propagation, feed forward, back propagation algorithm, weight, forecasting.


Keywords


return; neural network; back propagation; feed forward; back propagation algorithm; weight; forecasting.

Full Text:

PDF

Refbacks

  • There are currently no refbacks.



Creative Commons License
Jurnal Gaussian by Departemen Statistika Undip is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Flag Counter