From: Bitcoin price change and trend prediction through twitter sentiment and data volume
Study | Date range | Data days | Model-accuracy % | Data used |
---|---|---|---|---|
Predicting the price of Bitcoin using machine learning-2018 McNally et al. | 19/08/2013– 19/07/2016 | \(\sim\)1065 | LSTM-52.78% RNN-50.25% | Market |
An advanced CNN-LSTM model for cryptocurrency forecasting-2021 Livieris et al. | 01/01/2017–31/10/2020 | \(\sim\)1400 | (CNN-LSTM) Model1-55.03% Model2-53.64% MICDL-53.04% | Market |
Bitcoin price forecasting method based on CNN-LSTM hybrid neural network model-2019 Yan Li, Wei | 30/12/2016–31/08/2018 | \(\sim\)600 | (Precision) BP-59% CNN-64% LSTM-58% CNN-LSTM-64% | Market |
Bitcoin response to twitter sentiments Galenchuk et al. | 01/2014–09/2017 | \(\sim\)912 | RW-46.2% ARIMA-47.2% MLP-47.5% CNN-68.6% | Twitter market |
Price movement prediction of cryptocurrencies using sentiment analysis and machine learning- 2019 Valencia et al. | 16/02/2018– 21/04/2018 | \(\sim\)60 | MLP-72% SVM-55% RF - 44% | Twitter market |
Recurrent neural network based bitcoin price prediction by twitter sentiment analysis-2018 Pant et al. | 01/01/2018–30/06/2018 | \(\sim\)180 | RNN–77.62% | Twitter market |
Predicting bitcoin price fluctuation with Twitter sentiment analysis-2017 Stenqvist, Lönnö | 11/05/2017–11/06/2017 | \(\sim\)30 | No machine learning. Predicting direction solely on sentiment change in tweets 1hour_shift3-83.33% 30mins_shift4-78.78% 45mins_shift3-70.59% | Twitter Market |
Sentiment analysis based direction prediction in Bitcoin using deep learning algorithms and word embedding models-2020 Kilimci | 01/05/2019–01/08/2019 | \(\sim\)90 | GloVe-82.01% RNN-83.77% CNN-84.3% LSTM-87.45% FastText-89.13% | Twitter market |