Forecasting cryptocurrency prices time series using machine learning approach

Abstract

This paper describes the construction of the short-term forecasting model of cryptocurrencies’ prices using machine learning approach. The modified model of Binary Auto Regressive Tree (BART) is adapted from the standard models of regression trees and the data of the time series. BART combines the classic algorithm classification and regression trees (C&RT) and autoregressive models ARIMA. Using the BART model, we made a short-term forecast (from 5 to 30 days) for the 3 most capitalized cryptocurrencies: Bitcoin, Ethereum and Ripple. We found that the proposed approach was more accurate than the ARIMA-ARFIMA models in forecasting cryptocurrencies time series both in the periods of slow rising (falling) and in the periods of transition dynamics (change of trend).

Cryptocurrency price prediction using traditional statistical and machine‐learning techniques: A survey

“…Moreover, given the past dynamics and volatility trends, time-series models and autoregressive models are also adopted by researchers (e.g., Anupriya & Garg, 2018;Conrad et al, 2018;Derbentsev et al, 2019;Längkvist et al, 2014;Mohanty et al, 2018;Poyser, 2019;Roy et al, 2018;Troster et al, 2019).…”

“…This includes a collection of techniques, such as SVMs, ANNs, fuzzy logic, genetic algorithms, linear and nonlinear statistical models, DL and RL models, and so on (Atsalakis et al, 2019;Galeshchuk & Mukherjee, 2017;Hitam et al, 2019;Jiang & Liang, 2017;Längkvist et al, 2014;Lahmiri, 2011;Lahmiri & Bekiros, 2019;Nikou et al, 2019;Peng, Albuquerque, de Sá, Padula, & Montenegro, 2018;Radityo et al, 2017;Sarlin & Marghescu, 2011;Sin & Wang, 2017;Tupinambás, Cadence, & Lemos, 2018;Uras et al, 2020). schemes integrated various prediction models, including some of the popular classification techniques as well as some popular time-series forecasting techniques, while considering multiple aspects (Roy et al, 2018;Längkvist et al, 2014;Chakraborty & Roy, 2019;Derbentsev et al, 2019;Wang & Chen, 2020;Poyser, 2019). For example, some researchers studied the results using classical ARIMA models and different ML techniques, such as RF, linear discriminant analysis, logistic regression, and LSTM (Amjad & Shah, 2017;McNally et al, 2018;Saxena, Sukumar, Nadu, & Nadu, 2018).…”

“…A short‐term forecasting model was presented by Derbentsev, Datsenko, Stepanenko, and Bezkorovainyi (2019) to predict cryptocurrency prices of ripple, bitcoin, and ethereum using an ML approach. A binary autoregression tree was implemented in that paper that combines classification and ARIMA.…”

Penerapan Metode K-Means Dan C4.5 Untuk Prediksi Penderita Diabetes

“…Tahapan ini akan menguji validasi tingkat keakurasian hasil prediksi HbA1c terhadap metode K-Means dan C4. 5…”

“…Adapun penerapan proses metode K-Means dan C4 5. sebagai berikut: A. Proses K-Means Pada proses K-Means hanya akan dijelaskan sebagian.…”