Robert Kozub

EPPS 6323

Spring 2022

Project Proposal

Area/Topic

Predicting Bitcoin and Ethereum prices using historical/existing market data.

Research Statement

Research Question: Can data mining accurately predict the short-term price of bitcoin or Ethereum?

The issue wanting to be addressed with this project is to accurately predict the price of Bitcoin and Ethereum using knowledge mining and machine learning methods. What will the cost of these two cryptocurrencies be in the short-term? What will they be in the long term?

Since cryptocurrency has such high volatility, just like the stock market, it would be useful to know an accurate prediction to make better short-term investments. There is not as much research on predicting crypto price projects as the stock market due to its relatively new existence in the 21st century. Therefore, this project would provide more insight on if data mining can in fact be a useful tool for prediction.

Method

Using autocorrelation/linear regression models to compare previous prices and determine short-term price predictions. The data will be collected from an open-source site that stores the opening and closing price for Bitcoin and Ethereum over the last 5 years. In addition, comparing the correlation between Bitcoin and Ethereum using neural network classifiers will be performed.