Predictive Analytics in Cryptocurrency using Neural networks: A Comparative Study
Agha Salman Khan1, Peter Augustine2
1Agha Salman Khan, Department of Computer Science, CHRIST (Deemed to be University),Bangalore, Karnataka 560029,India.
2Dr. Peter Augustine, Department of Computer Science, CHRIST (Deemed to be University),Bangalore, Karnataka 560029,India.
Manuscript received on 13 March 2019 | Revised Manuscript received on 20 March 2019 | Manuscript published on 30 March 2019 | PP: 425-429 | Volume-7 Issue-6, March 2019 | Retrieval Number: F2350037619/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: This paper is concerned with assessing different neural network based predictive models. Each of these predictive models has one goal and that is to predict the price of a cryptocurrency, Bitcoin is the cryptocurrency taken into consideration. The models will be focusing on predicting the USD equivalent value of bitcoin using historical data and live data. The neural network models being assessed are a Convolutional Neural Network, and two variations of the Recurrent Neural Network that are Long Short Term Memory (LSTM) and Gated Recurrent Unit (GRU). The goal is to observe the validation loss of each model and also the time it takes to train or epoch for each training set which basically just determine its efficiency and performance. The results that are achieved are almost what was expected as LSTM outperforms CNN but the when we take a look at GRU, it is at par with LSTM.However, CNN is quicker at training or creating epochs and the validation loss is acceptable and not too high but it looks so when it is compared with the Recurrent Neural Networks such as Long Short Term Memory (LSTM) and Gated Recurrent Unit (GRU).
Keywords: Convolutional Neural Network (CNN), Cryptocurrency, Gated Recurrent Unit (GRU), Long Short Term Memory (LSTM)
Scope of the Article: Predictive Analysis