RSIS International

Cryptocurrency Price Prediction Using Long-Short Term Memory Model

Submission Deadline: 29th November 2024
November 2024 Issue : Publication Fee: 30$ USD Submit Now
Submission Deadline: 20th November 2024
Special Issue on Education & Public Health: Publication Fee: 30$ USD Submit Now
Submission Deadline: 05th December 2024
Special Issue on Economics, Management, Psychology, Sociology & Communication: Publication Fee: 30$ USD Submit Now

International Journal of Research and Scientific Innovation (IJRSI) | Volume V, Issue VII, July 2018 | ISSN 2321–2705

Cryptocurrency Price Prediction Using Long-Short Term Memory Model

Prashanth J R1, Vineetha S Das2

IJRISS Call for paper

1Research Scholar, Department of Mechanical Engineering
APJ Abdul Kalam Technological University, College of Engineering Trivandrum, Kerala, India
2Associate Professor, Department of Mechanical Engineering
APJ Abdul Kalam Technological University, College of Engineering Trivandrum, Kerala, India

Abstract- Cryptocurrency, is a decentralized digital or virtual currency. Use of cryptography for security makes it difficult to counterfeit. The primary cryptocurrency, the Bitcoin was launched in the year 2009 by Satoshi Nakamoto. Bitcoin, Ethereum, Ripple, Bitcoin cash, Bit connect, Dash, Ethereum Classic, Iota, Litecoin, Monero, Nem, Neo, Numeraire, Stratis, Waves etc. are some of the popular Cryptocurrencies. Cryptocurrencies started to gain attention in 2013 and since then witnessed a significant number of transactions and hence price fluctuations. The cryptocurrency market is just similar to stock market. It has gained public attention and so effective prediction of price movement of cryptocurrency will aid public to invest profitably in the system. This paper tries to predict the price of Cryptocurrencies. Machine learning techniques were implemented and the use of Adam optimizer and Long Short Term Memory (LSTM) network proved very efficient in predicting the prices of digital currencies.

Keywords: Cryptocurrency, price prediction, Machine Learning, Adam optimizer, LSTM

I. INTRODUCTION

Cryptocurrency, a technology supported virtual currency, is not issued by Government agency. It has highly volatile market price. The market capitalization of publicly traded cryptocurrenciеs is currently above $230 billion[1]. The digital currency came to existence with the introduction of Bitcoin in 2013. The cryptocurrency is not regulated by government agency. Peer-to-peer network of individuals process the transactions using internet. A block chain controls digital transactions [2]. Ethereum has the second-highest market capitalization and supports much more functionality than Bitcoin[1]. The Cryptocurrency market works similar to a stock market. Effective prediction of time series using conventional and various emerging techniques like machine learning enabled the market participants to earn promising rewards from stock markets. Application of such techniques to the market of Cryptocurrencies can help the investors to take advantage of liquidity of the system and thus gain substantial profit [3].