Author Information

Shital PandeyFollow

Is this project an undergraduate, graduate, or faculty project?

Graduate

Project Type

individual

Campus

Daytona Beach

Authors' Class Standing

Shital Pandey, Graduate

Lead Presenter's Name

Shital Pandey

Lead Presenter's College

DB College of Arts and Sciences

Faculty Mentor Name

Dr. Harihar Khanal

Abstract

The rise of cryptocurrencies as a novel asset class is a result of advancements in fintech, offering substantial research opportunities. Forecasting cryptocurrency prices is a complex task due to their inherent volatility and dynamic nature. In this study, we evaluate the performance of three recurrent neural network (RNN) algorithms—gated recurrent unit (GRU), long short-term memory (LSTM), and bidirectional LSTM (bi-LSTM)—in predicting the prices of Bitcoin (BTC), Ethereum (ETH), and Litecoin (LTC). The models' accuracy is assessed using the mean absolute percentage error (MAPE) metric. Our findings reveal that the GRU model surpasses the LSTM and bi-LSTM models in predicting the prices of all three cryptocurrencies, establishing it as the most effective algorithm. The GRU model achieves MAPE values of 0.2454% for BTC, 0.8267% for ETH, and 0.2116% for LTC. Conversely, the bi-LSTM model demonstrates the lowest prediction accuracy, with MAPE values of 5.990% for BTC, 6.85% for ETH, and 2.332% for LTC. In summary, the prediction models proposed in this study deliver accurate results that closely align with the actual prices of cryptocurrencies.

Did this research project receive funding support (Spark, SURF, Research Abroad, Student Internal Grants, Collaborative, Climbing, or Ignite Grants) from the Office of Undergraduate Research?

Yes, Student Internal Grants

Share

COinS
 

Cryptocurrency Price Prediction using Neural Networks and Deep Learning Techniques

The rise of cryptocurrencies as a novel asset class is a result of advancements in fintech, offering substantial research opportunities. Forecasting cryptocurrency prices is a complex task due to their inherent volatility and dynamic nature. In this study, we evaluate the performance of three recurrent neural network (RNN) algorithms—gated recurrent unit (GRU), long short-term memory (LSTM), and bidirectional LSTM (bi-LSTM)—in predicting the prices of Bitcoin (BTC), Ethereum (ETH), and Litecoin (LTC). The models' accuracy is assessed using the mean absolute percentage error (MAPE) metric. Our findings reveal that the GRU model surpasses the LSTM and bi-LSTM models in predicting the prices of all three cryptocurrencies, establishing it as the most effective algorithm. The GRU model achieves MAPE values of 0.2454% for BTC, 0.8267% for ETH, and 0.2116% for LTC. Conversely, the bi-LSTM model demonstrates the lowest prediction accuracy, with MAPE values of 5.990% for BTC, 6.85% for ETH, and 2.332% for LTC. In summary, the prediction models proposed in this study deliver accurate results that closely align with the actual prices of cryptocurrencies.

 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.