PREDIKSI HARGA BITCOIN MENGGUNAKAN ALGORITMA LONG SHORT TERM MEMORY (LSTM ) BERBASIS WEB

Cryptocurrency is a digital currency made from a series of codes or called blockchain, one of the cryptocurrencies is bitcoin. Prediction is a process that projects or imagines what might happen in the future based on past data or factors that affect the current situation. One technique is to use the LSTM (Long Short Term Memory) method. In this method the author uses quantitative methods. Quantitative research methods that use measurement and statistical analysis to collect, analyze, and interpret data. Long short term memory (LSTM) is a modeling developed from the Recurrent Neural Network (RNN) algorithm, a method designed to process sequence data, this model is useful for predicting bitcoin prices, and stock prices, using the sequence model the results of the validation loss value epoch 0 are above loss 0.01 at epoch 2 drop to below loss 0.01 with stable results until epoch 13, and the training loss value at epoch 0 is at loss 0. 06 at epoch 1 drops drastically to below loss 0.01 with stable results until epoch 13 is lower than the loss validation results.The level of accuracy of the presentation of the Long Short Term Memory LSTM method in predicting bitcoin prices, in the Root Mean Squared Error method for train data produces 17318.4049 and for test data produces 27921.84 and in the Mean Absolutet Percentage Error method for train data produces 3.24% and for test data produces 5.36%.
Keywords: Cryptocurrency, Blockchain, Bitcoin, lstm, rnn, data analysis