Are Bitcoins price predictable? Evidence from machine learning\n techniques using technical indicators

Abstract

The uncertainties in future Bitcoin price make it difficult to accurately\npredict the price of Bitcoin. Accurately predicting the price for Bitcoin is\ntherefore important for decision-making process of investors and market players\nin the cryptocurrency market. Using historical data from 01/01/2012 to\n16/08/2019, machine learning techniques (Generalized linear model via penalized\nmaximum likelihood, random forest, support vector regression with linear\nkernel, and stacking ensemble) were used to forecast the price of Bitcoin. The\nprediction models employed key and high dimensional technical indicators as the\npredictors. The performance of these techniques were evaluated using mean\nabsolute percentage error (MAPE), root mean square error (RMSE), mean absolute\nerror (MAE), and coefficient of determination (R-squared). The performance\nmetrics revealed that the stacking ensemble model with two base learner (random\nforest and generalized linear model via penalized maximum likelihood) and\nsupport vector regression with linear kernel as meta-learner was the optimal\nmodel for forecasting Bitcoin price. The MAPE, RMSE, MAE, and R-squared values\nfor the stacking ensemble model were 0.0191%, 15.5331 USD, 124.5508 USD, and\n0.9967 respectively. These values show a high degree of reliability in\npredicting the price of Bitcoin using the stacking ensemble model. Accurately\npredicting the future price of Bitcoin will yield significant returns for\ninvestors and market players in the cryptocurrency market.\n