Search

Bitcoin in the last decade, has become increasingly popular as a form of electronic currency. This paper aims to review the status and visual map position of research in the internationally Bitcoin studies indexed Scopus that used a bibliometric positioning overview. The research was carried out using bibliometric techniques. Data analysis as well as visualization utilizing VOS Viewer program and the Scopus function for analyze search results. In this review, the details collected applied to 3,281 documents issued from 2011 through 2019. The study reveals that Bouri, E. and Chinese Academy of Sciences, Studies were the most active individual scientists and affiliated institutions in Bitcoin studies. In bitcoin studies, the Computer Science and Lecture Notes in Computer Science Including Subseries Lecture Notes In Artificial Intelligence And Lecture Notes In Bioinformatics were the most areas of study and dissemination sources. There were one worldwide group maps with collaborative researchers. In order to identify the body of knowledge created from eight years of publication, this study constructed a convergence axis grouping comprising of bitcoin studies: Bitcoin, E-Finance, Blockchain, Privacy, Cybercrime, Computation, and Systems, abbreviated as the theme BFBPCCS. [ABSTRACT FROM AUTHOR]

"Predicting the future isn't magic, it's Artificial Intelligence". Cryptocurrency prices are much volatile, predicting the future prices of one such cryptocurrency prices "BITCOIN" using machine learning is the main objective of this study. The substantial part of this study is to predict the prices of bitcoin by using Long Short-Term Method (LSTM) on Recurrent Neural Network (RNN). The prices are predicted by training the network using the information of 4 years historical data of bitcoin prices. The time series data works well using Python and RNNs. The model works with different platforms in Python namely tensorflow and keras. The RNN remembers the inputs using the internal memory which makes it perfectly suited for solving machine learning problems involving sequential data. The model predicts the prices which are almost bordering on the actual prices. The subordinate parts involve comparing the model with another developed Automate Time Series (AutoTS) model. The results proclaim that both the models are best suited but LSTM model might give the results more precisely. [ABSTRACT FROM AUTHOR]