Crypto Currency Price Prediction on Ethereum Using Time Series Forecasting Models Arima and Facebook Prophet Models
Juvvala Sailaja, Kovvuri N Bhargavi, G. L. Narasamba Vanguri, Nagireddi Suryakala, Srinivasulu Thiruveedula · Advances in computer science research · 2024
Crypto currencies have emerged as a popular investment option in recent years, with Ethereum being one of the most prominent ones.Accurate price prediction of Ethereum can provide valuable insights to investors and traders for making informed decisions.In this study, we utilized two time series prediction models, ARIMA (Auto Regressive Integrated Moving Average) and Facebook Prophet, to predict the price of Ethereum.This research focuses on collecting legacy price data of Ethereum from a reliable source.The data was preprocessed to handle missing values and outliers.ARIMA and Facebook Prophet models were then implemented on the preprocessed data to generate Ethereum price forecasts.The models were trained using a time period of historical data and validated using a hold-out set of data.The MSE, which measures the squared discrepancies between predicted and real Ethereum prices, was used to assess the models' performance.Lower MSE values indicate better model performance.The results revealed that Facebook Prophet outperformed ARIMA in terms of MSE, indicating superior accuracy in Ethereum price prediction.The higher accuracy of Facebook Prophet may be attributed to it's ability to handle seasonality, trend changes, and outliers, which are common characteristics of crypto currency price data.In conclusion, this study demonstrates the effectiveness of time series forecasting models, specifically ARIMA and Facebook Prophet, in predicting Ethereum prices.The findings suggest that Facebook Prophet may be a more accurate model compared to ARIMA for Ethereum price prediction, as evidenced by lower MSE values.The study provides valuable insights for investors and traders interested in utilizing forecasting models for Ethereum price prediction, and may serve as a basis for further research in this area.