@phdthesis{digilib22236,
           month = {July},
           title = {JARINGAN SYARAF TIRUAN ALGORITMA BACKPROPAGATION UNTUK PERAMALAN HARGA INDEX SAHAM SYARIAH PADA BURSA EFEK INDONESIA},
          school = {UIN SUNAN KALIJAGA YOGYAKARTA},
          author = {NIM. 12650037 LINA NUR LATIFAH},
            year = {2016},
            note = {Nurochman, M.Kom},
        keywords = {Backpropagation, Forecasting, Neural Network},
             url = {https://digilib.uin-suka.ac.id/id/eprint/22236/},
        abstract = {Stocks are securities issued by a limited liability company or commonly known as the issuer. Stocks stated that the owner of these shares is also part-owner of the company. Thus, if an investor buys a stock, he became an owner or shareholder company. Knowing the stock prices in the future is a very important thing for investors. Forecasting process should seek the accuracy of forecasting results, although there will be a certain uncertainty of the company shares. So we need the best modeling for forecasting stock prices. Forecasting is done manually by the experience would be more difficult, therefore, forecasting the stock will more easily be done using the backpropagation algorithm.
One method of artificial neural network that can be applied in the case of data forecasting algorithms are Multilayer Backpropagation. Backpropagation through training, data that has been divided into training data and testing. Backpropagation network will be trained using the input data and targets that can then be simulated on test data. Steps taken to implement the backpropagation algorithm on data forecasting cases of data collection, creation of a network, set the value of the epoch, learning rate and momentum, training, up simulation test data.
Method Backpropagation Neural Network algorithm in this study were able to forecast the stock price data with the percentage of the truth of 95.9184\% of training data and test data at 84.5238\%, with a total of 280 data is the data, 196 training data and testing data 84.}
}