Deep Learning Based for Cryptocurrency Assistive System

Cryptocurrency is branded as a digital currency, an alternative exchange currency system with significant ramifications for the economies of rising nations and the global economy. In recent years, cryptocurrency has infiltrated almost all financial operations; hence, cryptocurrency trading is frequently recognized as one of the most popular and promising means of profitable investment. Lately, with the exponential growth of cryptocurrency investments, many Alternative Coins (Altcoins) resurfaced to mimic the fiat currency. There are several methods to forecast cryptocurrency prices that have been widely used in forecasting fiat and stock prices. Artificial Intelligence (AI) ,Machine Learning(ML) and Deep Learning(DL) provide a different perspective on how investors can estimate crypto price trend and movement. In this paper, as cryptocurrency price is time-dependent, Recurrent Neural Network (RNN) is presented due to RNN’s nature, which is well suited for Time Series Analysis (TSA). The topology of the proposed RNN model consists of three stages which are model groundwork, model development, and testing and optimization. The RNN architecture is extended to three different models specifically Long Short-Term Memory (LSTM), Gat-ed Recurrent Unit (GRU), and Bi-Directional Long Short-Term Memory (LSTM). There are a few hyperparameters that affect the accuracy of the deep learning model in predicting cryptocurrency prices. Hyperparameter tuning set the basis for optimizing the model to improve the accuracy of cryptocurrency prediction. Next, the models were tested with data from different coins listed in the cryptocurrency market. Then, the model was experiment-ed with different input features to figure out how accurate and robust these models in predicting the cryptocurrency price. GRU has the best accuracy in forecasting the cryptocurrency prices based on the values of Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and Executional Time, scoring 2.2201, 0.8076, and 200s using the intraday trading strategy as input features