The goal of our work was to select a neural network architecture that would give the best prediction of the Bitcoin exchange rate using historical data. Our work fits into the very important topic of predicting the value of the cryptocurrency exchange rate, and makes use of recent data which, as a result of the high Bitcoin exchange rate dynamics of the last year, differs significantly from those of previous years. We propose and test a number of neural network-based architectures and conduct a discussion of the results. Unlike previous state-of-the-art works, we conducted a comprehensive comparison of three different neural network-based models: MLP (multilayer perceptron), LSTM (long short-term memory) and CNN (convolutional neural network). We tested them for a wide range of parameters. The results we present are, to the best of our knowledge, the most up to date when it comes to the application of artificial intelligence methods for the prediction of cryptocurrency exchange rates. The best-performing architectures were used for a website that gives real-time predictions of the Bitcoin exchange rate. The website is available at http://stpbtc-ii.up.krakow.pl/. Source codes of our research are available to download in order to make our experiment reproducible.
Akif Akgül, Eyyüp Ensari Şahin, Fatma Yıldız ŞENOL
Crypto assets succeeded in making their name known to large masses with Bitcoin, which emerged as a result of the creation of the first genesis block in 2008. Until 2010, the aforementioned recognition showed itself mostly in areas such as games, but over time it managed to enter the portfolios of individual investors. Especially as of end of 2017, the rapid increases in monetary value quickly attracted the attention of corporate companies and then the (Central Banks). These assets have created different alternatives (also know as altcoins) by working and have managed to become one of the important financial instruments today. This study has examined in detail the techniques (Chaos theory, Onchain analysis and Sentiment analysis) developed on the price predictions of crypto assets, which are very important in terms of the number and quality of investors. In the study, findings were obtained that new techniques such as onchain and sentiment are more prominent in estimating crypto asset prices compared to traditional asset price estimation methods of crypto assets and that these techniques can make consistent estimations.
Abstract Cryptocurrency investment especially Bitcoin, has become favourable over recent years due to promising returns in the future. However, the movement of price is mainly speculation-based as these currencies are still new in the market. The COVID-19 outbreak boosted research interest in predicting the price fluctuation of Bitcoin since cryptocurrency trading produced many millionaires. Five lexicon-based Twitter sentiment analysis approaches are examined to see the effect of Tweets on the price of Bitcoin during the pre- and post- COVID-19 period. Results show that negative Twitter sentiments affected the price of Bitcoin pre- COVID-19 and the second year of post- COVID-19 when Elon Musk actively criticised Bitcoin on Twitter.
Since bitcoin has gained recognition as a valuable asset, researchers have begun to use machine learning to predict bitcoin price. However, because of the impractical cost of hyperparameter optimization, it is greatly challenging to make accurate predictions. In this paper, we analyze the prediction performance trends under various hyperparameter configurations to help them identify the optimal hyperparameter combination with little effort. We employ two datasets which have different time periods with the same bitcoin price to analyze the prediction performance based on the similarity between the data used for learning and future data. With them, we measure the loss rates between predicted values and real price by adjusting the values of three representative hyperparameters. Through the analysis, we show that distinct hyperparameter configurations are needed for a high prediction accuracy according to the similarity between the data used for learning and the future data. Based on the result, we propose a direction for the hyperparameter optimization of the bitcoin price prediction showing a high accuracy.
Bitcoin is yet to be assumed as a worthy cryptocurrency and rewarding asset in the global market. As polynomial-based neural networks (PBNNs) are very robust and more accurate in modeling stock price prediction, their advantage in Bitcoin pricing needs to be analyzed. In this study, the robustness of PBNNs, based on Chebyshev (CPBNN) and Legendre (LPBNN), is blended with the proposed algorithm, coined as the mutated climb monkey algorithm (MCMA), to control the estimation of network parameters to accurately predict the one-day-ahead Bitcoin price. The performance was evaluated by a comparative analysis of the testing of both CPBNN and LPBNN with each of the six algorithms under consideration on three different datasets collected within the same time interval. As the use of a few evaluation criteria will not be able to identify an efficient predictor model, this study also proposes the use of a Multi-Criteria Decision-Making (MCDM) framework to rank all models using 15 different evaluation criteria. The ranking of the models clearly indicates that the proposed MCMA algorithm outperforms all other algorithms under study. The convergence plots of the top two models for the datasets also indicate that the PBNN using MCMA for learning predicts better results.
Cryptocurrency is the most secure, traceable, and reliable intangible currency because it uses blockchain technology. It eliminates the middle layer of financial institutes in the traditional economic system. Because of high returns in cryptocurrencies, investors and other firms invest a lot of money. But the prices of the cryptocurrencies are uncertain. Prices of cryptocurrencies are influenced by many factors like sentiments, trading volume, and similar. Researchers are doing plenty of work to predict the accurate prices of various cryptocurrencies. However, many of these methods cannot be used in real-time. Several deep learning models such as Neural networks (NN), Long short-term memory (LSTM), and Gated recurrent unit (GRU) have been utilized by researchers for predicting the price of cryptocurrencies and yet, are unable to achieve significant results. This work combines LSTM and GRU with sentiment analysis to precisely estimate bitcoin values. We have used Root means square error (RMSE) to evaluate the model performance with and without sentiments. Empirically, we have compared the results with the other state-of-the-art models and found better results using the proposed hybrid model incorporated with sentiments.
The total capital in cryptocurrency markets is around two trillion dollars in 2022, which is almost the same as Apple’s market capitalisation at the same time. Increasingly, cryptocurrencies have become established in financial markets with an enormous number of transactions and trades happening every day. Similar to other financial systems, price prediction is one of the main challenges in cryptocurrency trading. Therefore, the application of artificial intelligence, as one of the tools of prediction, has emerged as a recently popular subject of investigation in the cryptocurrency domain. Since machine learning models, as opposed to traditional financial models, demonstrate satisfactory performance in quantitative finance, they seem ideal for coping with the price prediction problem in the complex and volatile cryptocurrency market. There have been several studies that have focused on applying machine learning for price and movement prediction and portfolio management in cryptocurrency markets, though these methods and models are in their early stages. This survey paper aims to review the current research trends in applications of supervised and reinforcement learning models in cryptocurrency price prediction. This study also highlights potential research gaps and possible areas for improvement. In addition, it emphasises potential challenges and research directions that will be of interest in the artificial intelligence and machine learning communities focusing on cryptocurrencies.
As the cryptocurrency trading market has grown significantly in recent years, the number of comments related to cryptocurrency has increased tremendously in social media platforms. Due to this, sentiment analysis of the cryptocurrency-related comments has become highly desirable to give a comprehensive picture of peoples’ opinions about the trend of the market. In this regard, we perform cryptocurrency-related text sentiment classification using tweets based on positive and negative sentiments. For increasing the efficacy of the sentiment analysis, we introduce a novel deep neural network hybrid architecture which is composed of an embedding layer, a convolution layer, a group-wise enhancement mechanism, a bidirectional layer, an attention mechanism, and a fully connected layer. Local features are derived using a convolution layer, and weight values associated with intuitive features are developed using the group-wise enhancement mechanism. After feeding the improved context vector to the bidirectional layer to grab global features, the attention mechanism and the fully connected layer have been employed. The experimental findings indicate that the proposed architecture outperforms the state-of-the-art architectures with an accuracy value of 93.77%.
Haruna Umar Yahaya, John Sunday Oyinloye, Samuel Olorunfemi Adams
The future of e-money is crypocurrencies, it is the decentralize digital and virtual currency that is secured by cryptography. It has become increasingly popular in recent years attracting the attention of the individual, investor, media, academia and governments worldwide. This study aims to model and forecast the volatilities and returns of three top cryptocurrencies, namely; Bitcoin, Ethereum and Binance Coin. The data utilized in the study was extracted from the higher market capitalization at 31st December, 2021 and the data for the period starting from 9th November, 2017 to 31st December 2021. The Generalised Autoregressive conditional heteroscedasticity (GARCH) type models with several distributions were fitted to the three cryptocurrencies dataset with their performances assessed using some model criteria. The result shows that the mean of all the returns are positive indicating the fact that the price of this three crptocurrencies increase throughout the period of study. The ARCH-LM test shows that there is no ARCH effect in volatility of Bitcoin and Ethereum but present in Binance Coin. The GARCH model was fitted on Binance Coin, the AIC and log L shows that the CGARCH is the best model for Binance Coin. Automatic forecasting was perform based on the selected ARIMA (2,0,1), ARIMA (0,1,2) and the random walk model which has the lowest AIC for ETH-USD, BNB-USD and BTC-USD respectively. This finding could aid investors in determining a cryptocurrency's unique risk-reward characteristics. The study contributes to a better deployment of investor’s resources and prediction of the future prices the three cryptocurrencies.
In recent years, digital currencies have flourished on a considerable scale, and the markets of digital currencies have generated a nonnegligible impact on the whole financial system. Under this background, the accurate prediction of cryptocurrency prices could be a prerequisite for managing the risk of both cryptocurrency markets and financial systems. Considering the multiscale attributes of cryptocurrency price, we match the different machine learning algorithms to corresponding multiscale components and construct the ensemble prediction models based on machine learning and multiscale analysis. The Bitcoin price series, respectively, from 2017/11/24 to 2020/4/21 and 2020/4/22 to 2020/11/27, is selected as the training and prediction datasets. The empirical results show that the ensemble models can achieve a prediction accuracy of 95.12%, with better performance than the benchmark models, and the proposed models are robust in upward and downward market conditions. Meanwhile, the different algorithms are applicable for components with varying time scales.
Azeez A. Oyedele, Anuoluwapo Ajayi, Lukumon O. Oyedele, Sururah A. Bello · 5 authors
The emergence of cryptocurrencies has drawn significant investment capital in recent years with an exponential increase in market capitalization and trade volume. However, the cryptocurrency market is highly volatile and burdened with substantial heterogeneous datasets characterized by complex interactions between predictors, which may be difficult for conventional techniques to achieve optimal results. In addition, volatility significantly impacts investment decisions; thus, investors are confronted with how to determine the price and assess their financial investment risks reasonably. This study investigates the performance evaluation of a genetic algorithm tuned Deep Learning (DL) and boosted tree-based techniques to predict several cryptocurrencies' closing prices. The DL models include Convolutional Neural Networks (CNN), Deep Forward Neural Networks, and Gated Recurrent Units. The study assesses the performance of the DL models with boosted tree-based models on six cryptocurrency datasets from multiple data sources using relevant performance metrics. The results reveal that the CNN model has the least mean average percentage error of 0.08 and produces a consistent and highest explained variance score of 0.96 (on average) compared to other models. Hence, CNN is more reliable with limited training data and easily generalizable for predicting several cryptocurrencies' daily closing prices. Also, the results will help practitioners obtain a better understanding of crypto market challenges and offer practical strategies to lower risks.
Our work presents two fundamental contributions. On the application side, we tackle the challenging problem of predicting day-ahead crypto-currency prices. On the methodological side, a new dynamical modeling approach is proposed. Our approach keeps the probabilistic formulation of the state-space model, which provides uncertainty quantification on the estimates, and the function approximation ability of deep neural networks. We call the proposed approach the deep state-space model. The experiments are carried out on established cryptocurrencies (obtained from Yahoo Finance). The goal of the work has been to predict the price for the next day. Benchmarking has been done with both state-of-the-art and classical dynamical modeling techniques. Results show that the proposed approach yields the best overall results in terms of accuracy.
This article proposes a graph neural network strategy (GNN), in which the long short-term memory (LSTM) and graph convolution network (GCN) are applied to capture both temporal and spatial features to forecast the price of Bitcoin, Litecoin, Ethereum, and Dash Coin with the ‘stable-coin’ Tether (USDT) and financial stress index (FSI). The main results show that the GNN strategy has better performance than univariate LSTM and multivariate LSTM in all of the seven steps forward forecasting. A sensitivity check shows that USDT and FSI/sub-FSI are important factors in the construction of the graphs and they verify the validity of the results.
This study uses the API of Upbit, one of Korea’s cryptocurrency exchanges, to predict continuous time series for a limited period and cryptocurrencies using LSTM, a machine learning technique. The trading (buying and selling) point algorithm presented in this study was used to conduct experimental research on efficient profit creation for cryptocurrency investment. Several related studies have shown the results of time series prediction for long-term forecasts, such as a week or several months. Still, they have not attempted to make an ultra-short-term prediction in units of one minute. This paper attempts such a 1 min prediction. This is an experiment to create efficient profits by setting efficient trading (buying and selling) points using machine learning techniques and repeating these operations by an algorithm. Applying it to cryptocurrency shows the possibility of time series prediction.
David Opeoluwa Oyewola, Emmanuel Gbenga Dada, Juliana Ngozi Ndunagu
Cryptocurrency is an advanced digital currency that is secured by encryption, making it nearly impossible to forge or duplicate. Many cryptocurrencies are blockchain-based with decentralized networks. The prediction of cryptocurrency prices is a very difficult task because of the absence of an appropriate analytical basis to substantiate their claims. Cryptocurrencies are also dependent on several variables, such as technical advancement, internal competition, market pressure, economic concerns, security, and political considerations. This paper proposed the hybrid walk-forward ensemble optimization technique and applied it to predict the daily prices of fifteen cryptocurrencies, such as Cardano (ADA-USD), Bitcoin (BTC-USD), Dogecoin (DOGE-USD), Ethereum Classic (ETC-USD), Chainlink (LINK-USD), Litecoin (LTC-USD), NEO (NEO-USD), Tron (TRX-USD), Tether (USDT-USD), NEM (XEM-USD), Stellar (XLM-USD), Ripple (XRP-USD), and Tezos (XTZ-USD). A performance comparison of these cryptocurrencies was done using classical statistical models, machine learning algorithms, and deep learning algorithms on different cryptocurrency time series. Simulation results show that our proposed model performed better in terms of cryptocurrency prediction accuracy compared to the classical statistical model and machine and deep learning algorithms used in this paper.
We employ and analyze various machine learning models for daily cryptocurrency market prediction and trading. We train the models to predict binary relative daily market movements of the 100 largest cryptocurrencies. Our results show that all employed models make statistically viable predictions, whereby the average accuracy values calculated on all cryptocurrencies range from 52.9% to 54.1%. These accuracy values increase to a range from 57.5% to 59.5% when calculated on the subset of predictions with the 10% highest model confidences per class and day. We find that a long-short portfolio strategy based on the predictions of the employed LSTM and GRU ensemble models yields an annualized out-of-sample Sharpe ratio after transaction costs of 3.23 and 3.12, respectively. In comparison, the buy-and-hold benchmark market portfolio strategy only yields a Sharpe ratio of 1.33. These results indicate a challenge to weak form cryptocurrency market efficiency, albeit the influence of certain limits to arbitrage cannot be entirely ruled out.
The rising of bitcoin’s user as a digital currency and investments causing an instability and an uncertainty in price movement and increasing the risk of trading, therefore in this study we try to forecast the future value of bitcoin price using ARIMA Models. 2 candidate models are selected by the lowest value of AIC and using the performance indicators ME, RSME, MAE, MPE, and MAPE conclude ARIMA (1,1,0) are the best ARIMA model, then the next 5 months future price forecasted using the best model. While ARIMA (1,1,0) is the best model, the model failed to follow price movement as shown in the forecasted price.
Virtual currencies have been declared as one of the financial assets that are widely recognized as exchange currencies. The cryptocurrency trades caught the attention of investors as cryptocurrencies can be considered as highly profitable investments. To optimize the profit of the cryptocurrency investments, accurate price prediction is essential. In view of the fact that the price prediction is a time series task, a hybrid deep learning model is proposed to predict the future price of the cryptocurrency. The hybrid model integrates a 1-dimensional convolutional neural network and stacked gated recurrent unit (1DCNN-GRU). Given the cryptocurrency price data over the time, the 1-dimensional convolutional neural network encodes the data into a high-level discriminative representation. Subsequently, the stacked gated recurrent unit captures the long-range dependencies of the representation. The proposed hybrid model was evaluated on three different cryptocurrency datasets, namely Bitcoin, Ethereum, and Ripple. Experimental results demonstrated that the proposed 1DCNN-GRU model outperformed the existing methods with the lowest RMSE values of 43.933 on the Bitcoin dataset, 3.511 on the Ethereum dataset, and 0.00128 on the Ripple dataset.
Cryptocurrencies emerged with the invention of Bitcoin in 2008 and in a short time they managed to attract the attention of a significant number of investors. Since 2012, cryptocurrency markets have become markets where unpredictable transaction volumes take place around the world. Bitcoin is the most recognized asset which has the highest trading volume in cryptocurrency markets today. This is due to the dominance effect of Bitcoin on cryptocurrency markets. It is a matter of curiosity in the world and in our country whether savings are transferred from traditional financial markets to cryptocurrency markets. In this study, the possible relationships between Bitcoin trading volumes and Bitcoin price changes between 2017 and 2021, and the trading volumes realized in Borsa Istanbul (BIST) and the BIST 100 index were examined with the Granger causality approach and the direction of these relationships was tried to be determined using the Toda Yamamoto causality model. As a result of the examination, it has been determined that there is no causal relationship between Bitcoin trading volumes and Borsa Istanbul trading volumes according to the Toda Yamamoto approach.
<span lang="EN-US">Cryptocurrency is a virtual or digital currency used in financial systems that utilizes blockchain technology and cryptographic functions to gain transparency, decentralization, and conservation. Cryptocurrency prices have a high level of fluctuation; thus, tools are needed to monitor and predict them. RNN is a deep learning model that is capable of strongly predicting data time series. Some types of Recurrent Nureal Network layers, such as Long Short Term Memory, have been used in previous studies to prediction common used currency. In this study, we used the Gate Recurrent Unit and Bidirectional</span><span lang="EN-US">–</span><span lang="EN-US">LSTM hybrid model to predict cryptocurrency prices to improve the accuracy of previously proposed prediction LSTM Model to predict the Bitcoin, Using four cryptocurrencies (Bitcoin, Ehtereum, Ripple, and Binance), we obtained very good results with RMSE after normalization the results get closer to 0 and with MAPE values all below &lt;10%.</span>
Çalışmanın temel amacı; son on yıla damga vuran kripto paralar arasındaki getiri ve risk açısından nedensellik ilişkisini tespit etmektir. Bu amaçla piyasa değeri en yüksek 10 kripto paradan en fazla verisi bulunan Bitcoin, Ethereum, Litecoin, Stellar, Ripple arasında 10 model oluşturulmuş ve her model, Granger nedensellik ve Hafner-Herwatz varyansta nedensellik testleri test edilmiştir. Çalışmada 23 Şubat 2017 ile 18 Haziran 2021 tarihleri arasındaki günlük verilerden oluşan 1577 gözlem kullanılmıştır. Nedensellik analizi sonuçlarına göre seçili kripto paralar arasında ortalamada Ethereum - Litecoin hariç diğer değişkenler arasında Granger nedensellik ilişkisi, varyansta ise Bitcoin - Ethereum ve Bitcoin - Litecoin hariç diğer değişkenler arasında varyansta nedensellik ilişkisi tespit edilmiştir.
This study analysis forecasting the bitcoin exchange rate against the USD. The dataset selected for this study starts from January 2015 to June 2022. This study's methodology uses autoregressive integrated moving average forecasting (ARIMA). The overall outcomes of this study were gathered from the statistical software Minitab 21.1. The Box Jenkins approaches are also used to predict the best model. To determine the ARIMA model parameter, this study did autocorrelation function (ACF) and partial autocorrelation function (PACF) analyses. According to the Box-Cox transformation method, log transformation was selected. The outcome demonstrates that the seasonal with the regular difference in the Bitcoin exchange rate against the USD is a stationary data series. The forecasting model used in this study is ARIMA (1,1,0) (2,1,1)12. This predicted model is identified through the Mean squared error by comparing the other guessing ARIMA models. After the prediction, 5 Month bitcoin exchange rate against the USD. Investors will be able to estimate the bitcoin exchange rate against the USD with the use of this information, but volatility must also be properly watched. This will aid investors in making better investment decisions and increase profits. In future studies, better consider another exchange rate of BTC and software experts will develop such type of software based on ARIMA models for prediction.