Bitcoin Prediction Analysis Using Deep Learning Techniques

Bitcoin is the first and most commonly used digital cryptocurrency in the world. It is also used for electronic transaction though it does not exist physically like hard notes. For investors, it is now been regarded as an investment opportunity. As it is volatile highly in nature, therefore it is required to have good prediction algorithms. On the basis of those algorithms, one can make investment decisions. Cryptocurrency is expanding and getting more attention among people, therefore, a precise prediction of bitcoin price is becoming an important feature in the digital financial market, although there are many studies that have leveraged machine learning for more accurate prediction for bitcoin in which different data structures and data features are adopted. In this paper, we have used the daily price which is the closed price. A comparison has been made between LSTM and deep hybrid neural network (a combination of LSTM and CNN). We found that LSTM outperforms than other algorithms.