Cryptocurrencies are a digital way of money in which all transactions are held electronically. It is a soft currency which doesn’t exist in the form of hard notes physically. Here, we are emphasizing the difference of fiat currency which is decentralized that without any third-party intervention all virtual currency users can get the services. However, getting services of these cryptocurrencies impacts on international relations and trade, due to its high price volatility. There are several virtual currencies such as bit-coin, ripple, ethereum, ethereum classic, lite coin, etc. In our study, we especially focused on a popular cryptocurrency, i.e., bitcoin. From many types of virtual currencies, bitcoin has a great acceptance by different bodies such as investors, researchers, traders, and policy-makers. To the best of our knowledge, our target is to implement the efficient deep learning-based prediction models. Specifically long short-term memory (LSTM) and gated recurrent unit (GRU) to handle the price volatility of bitcoin and to obtain high accuracy. Our study involves comparing these two time series deep learning techniques and proved the efficacy in forecasting the price of bitcoin.
Non-Fungible Tokens (NFTs) represent deeds of ownership, based on blockchain technologies and smart contracts, of unique crypto assets on digital art forms (e.g., artworks or collectibles). In the spotlight after skyrocketing in 2021, NFTs have attracted the attention of crypto enthusiasts and investors intent on placing promising investments in this profitable market. However, the NFT financial performance prediction has not been widely explored to date. In this work, we address the above problem based on the hypothesis that NFT images and their textual descriptions are essential proxies to predict the NFT selling prices. To this purpose, we propose MERLIN, a novel multimodal deep learning framework designed to train Transformer-based language and visual models, along with graph neural network models, on collections of NFTs' images and texts. A key aspect in MERLIN is its independence on financial features, as it exploits only the primary data a user interested in NFT trading would like to deal with, i.e., NFT images and textual descriptions. By learning dense representations of such data, a price-category classification task is performed by MERLIN models, which can also be tuned according to user preferences in the inference phase to mimic different risk-return investment profiles. Experimental evaluation on a publicly available dataset has shown that MERLIN models achieve significant performances according to several financial assessment criteria, fostering profitable investments, and also beating baseline machine-learning classifiers based on financial features.
Marek Zatwarnicki, Krzysztof Zatwarnicki, Piotr Stolarski
In 2020 and 2021, the cryptocurrency market attracted millions of new traders and investors. Lack of regulation, high liquidity, and modern exchanges significantly lowered the entry threshold for new market participants. In 2021, over 5 million Americans were regularly involved in cryptocurrency trading. At that time, the interest in market indicators and trading strategies remained low, leading to the conclusion that most investors did not use decision-support indicators. The correct and backtested use of technical analysis signals can give the trader a significant advantage over most market participants. This work introduces an algorithmic approach to examining the effectiveness of the signals generated by one of the most popular market indicators, the Relative Strength Index (RSI). A model corresponding to an actual cryptocurrency exchange was used to backtest the strategies. The results show that the RSI as a momentum indicator in the cryptocurrency market involves high risk. Using alternative RSI applications can allow traders to gain an advantage in the cryptocurrency market. Comparing the results with the traditional buy and hold strategy shows the credible potential of the indicated method and the usage of signals generated by the technical analysis indicators.
Kate Murray, Andrea Rossi, Diego Carraro, Andrea Visentin
Traders and investors are interested in accurately predicting cryptocurrency prices to increase returns and minimize risk. However, due to their uncertainty, volatility, and dynamism, forecasting crypto prices is a challenging time series analysis task. Researchers have proposed predictors based on statistical, machine learning (ML), and deep learning (DL) approaches, but the literature is limited. Indeed, it is narrow because it focuses on predicting only the prices of the few most famous cryptos. In addition, it is scattered because it compares different models on different cryptos inconsistently, and it lacks generality because solutions are overly complex and hard to reproduce in practice. The main goal of this paper is to provide a comparison framework that overcomes these limitations. We use this framework to run extensive experiments where we compare the performances of widely used statistical, ML, and DL approaches in the literature for predicting the price of five popular cryptocurrencies, i.e., XRP, Bitcoin (BTC), Litecoin (LTC), Ethereum (ETH), and Monero (XMR). To the best of our knowledge, we are also the first to propose using the temporal fusion transformer (TFT) on this task. Moreover, we extend our investigation to hybrid models and ensembles to assess whether combining single models boosts prediction accuracy. Our evaluation shows that DL approaches are the best predictors, particularly the LSTM, and this is consistently true across all the cryptos examined. LSTM reaches an average RMSE of 0.0222 and MAE of 0.0173, respectively, 2.7% and 1.7% better than the second-best model. To ensure reproducibility and stimulate future research contribution, we share the dataset and the code of the experiments.
We considered the daily price dynamics of the US Bitcoin market in the period from 2015 to 2022. In the first step, we used a singular value decomposition (SVD) entropy method for assessing time-varying informational efficiency over different time scales, from weeks to quarters. It was shown that the US Bitcoin market has been informationally efficient most of the time, except for some isolated periods where the returns exhibited deviations from the random behavior. The COVID-19 pandemic has not impacted the informational efficiency. This suggests that the Bitcoin market is unpredictable, and no reliable predictions can be obtained. A further analysis was carried out by considering the recurrence intervals for different positive and negative returns. We found that the distribution of recurrence intervals for positive and negative returns is asymmetric, with mean values higher for negative returns. We found that the distribution of recurrence intervals can be described by a stretching exponential distribution, such that the empirical and analytical hazard probabilities as functions of the elapsed time show good agreement.
Numerous research have been done to predict cryptocurrency prices since cryptocurrency prices affect global economic and monetary systems. However, investigations using linear connection approaches and technical analysis indicators frequently fall short of providing an explanation for changes in the pattern of BitCoin pricing. This paper is proposed to study time-varying parameters with long short-term memory (LSTM). The study is investigated on a dataset retrieved from Binance from March 2022 to April 2022. The proposed LSTM used a variety of hyperparameter settings, particularly time parameters, to predict the cryptocurrency price (BTC/USDT) on the dataset. Additionally, it is evaluated in terms of mean absolute percentage error (MAPE) in comparison to smooth moving average (SMA), weighted moving average (WMA), and exponential moving averages (EMA). From the investigation, using the previous 3 days for prediction gives the lowest of the MAPE values and the proposed LSTM outperformed the other models. When considering the last three days' value of pricing, the indicated LSTM offers the best accurate prediction, with a MAPE percentage of 0.0927%.
Irfan Abdul Karim Shaikh, P.Vamsi Krishna, Swagat Gourav Biswal, A. Sasi Kumar · 6 authors
Digital currency is a way of currency utilized in the digital world namely electronic devices or digital forms. Many terms are alternative words for digital currency such as cyber cash, digital money, and electronic money. Cryptocurrency is a type of asset that has developed due to the progression of financial technology and it has made a tremendous chance for research workers. Cryptocurrency price prediction is challenging because of the dynamism and price volatility. The electronic economy is severely hazardous and should be advanced with greater caution, to minimize or avoid the risk that occurs in this case. Therefore, this study develops a new Bayesian optimization with Stacked Sparse Autoencoder based Cryptocurrency Price Prediction (BOSSAE-CPP) model. The major intention of the BOSSALCPP technique lies in the effectual prediction of cryptocurrency prices. To attain this, the BOSSAE-CPP technique exploits SSAE model for price prediction process. Moreover, the BO technique is used to optimally choose the hyperparameter values of the SSAE model and results in enhanced predictive outcomes. To deliberate the enhanced outcomes of the BOSSALCPP technique, extensive experimentation study is made. The comparison study highlighted the improved performance of the BOSSAE-CPP technique.
Krishna Pal Sharma, Shivam Kumar Singh, Ankur Choudhary, Himanshu Goel
The world's most well-known cryptocurrency, Bitcoin, is much attractive to financial market players due to its recent price boom and fall. It is very difficult to anticipate because of the substantial volatility of the Bitcoin conversion scale. Forecasting of its behavior is crucial for the financial industry sectors. Some of the research works have been used machine learning techniques on past trends data of Bitcoin for its price predictions but they lack in accuracy. Therefore, the aim of this paper is to look into the global crypto-currency price movement trends about social media communication data as well as historical data trends. The thought is to analyze informative trends in online groups and social media networks to fully understand and extract useful information which could be used to improve the prediction accuracy of cryptocurrency price fluctuations. In this work we examine the behavior of Artificial Neural Networks (ANN), Recurrent Neural Networks (RNN) with LSTM and GRU, ARIMA model, Random forests(RF), and others for price direction predictions. Similarly, the SVM and ANN are used to predict Bitcoin’s minimum, maximum, and closing prices. The outcomes of the forecasts are also utilized as inputs to improve forecasts of price movement. The results showed that performance was greatly improved by the characteristics that were picked as well as by the effective machine learning and deep learning techniques.
Ashikur Rahaman, Abu Kowshir Bitto, Khalid Been Md. Badruzzaman Biplob, Md. Hasan Imam Bijoy · 6 authors
As opposed to other fiat currencies, bitcoin has no relationship with banks. Its price fluctuation is largely influenced by fresh blocks, news, mining information, support or resistance levels, and public opinion. Therefore, a machine-learning model will be fantastic if it learns from data and tells or indicates if we need to purchase or sell for a little period. In this study, we attempted to create a tool or indicator that can gather tweets in real-time using tweepy and the Twitter application programming interface (API) and report the sentiment at the time. Using the renowned Python module "FBProphet," we developed a model in the second phase that can gather historical price data for the bitcoin to US dollar (BTCUSD) pair and project the price of bitcoin. In order to provide guidance for an intelligent forex trader, we finally merged all of the models into one form. We traded with various models for a very little number of days to validate our bitcoin trading indicator (BTI), and we discovered that the combined version of this tool is more profitable. With the combined version of the instrument, we quickly and with little error root mean square error (RMSE: 1,480.58) generated a profit of $1,000.71 USD.
Abstract Mean-variance portfolio optimization models are sensitive to uncertainty in risk-return estimates, which may result in poor out-of-sample performance. In particular, the estimates may suffer when the number of assets considered is high and the length of the return time series is not sufficiently long. This is precisely the case in the cryptocurrency market, where there are hundreds of crypto assets that have been traded for a few years. We propose enhancing the mean-variance (MV) model with a pre-selection stage that uses a prototype-based clustering algorithm to reduce the number of crypto assets considered at each investment period. In the pre-selection stage, we run a prototype-based clustering algorithm where the assets are described by variables representing the profit-risk duality. The prototypes of the clustering partition are automatically examined and the one that best suits our risk-aversion preference is selected. We then run the MV portfolio optimization with the crypto assets of the selected cluster. The proposed approach is tested for a period of 17 months in the whole cryptocurrency market and two selections of the cryptocurrencies with the higher market capitalization (175 and 250 cryptos). We compare the results against three methods applied to the whole market: classic MV, risk parity, and hierarchical risk parity methods. We also compare our results with those from investing in the market index . The simulation results generally favor our proposal in terms of profit and risk-profit financial indicators. This result reaffirms the convenience of using machine learning methods to guide financial investments in complex and highly-volatile environments such as the cryptocurrency market.
As the most-traded digital asset, Bitcoin receives a tremendous increase in investment interests. The primary purpose of this paper is to examine and compare two commonly applied machine learning algorithms on their capability and feasibility in Bitcoin price prediction. The regression models of Extreme Gradient Boosting and Long Short-Term Memory are selected as the investigated objects in this paper. The experiments are evaluated by considering the extra impacts of sample dimensions and time-series frequency, along with the trade-offs between prediction accuracy (measured by residual error) and computational efficiency (measured by computing time). Long Short-Term Memory, as a more convoluted deep learning method, achieves better accuracy when a daily dataset with limited input features is used. However, its predictability has considerably decreased and thus become less efficient than XGBoost, when more peripheral information and higher frequency data points (trading price every 15 minutes) are available. Besides algorithmic complexity, Long Short-Term Memory also takes much longer computing time than Extreme Gradient Boosting, making it a less applicable model to use when dealing with large sample sizes.
The research purpose of this paper is to obtain an algorithm model with high prediction accuracy for the price of Bitcoin on the next day through random forest regression and LSTM, and to explain which variables have influence on the price of Bitcoin. There is much prior literature on Bitcoin price prediction research, and the research methods mainly revolve around the ARMA model of time series and the LSTM algorithm of deep learning. Although it cannot be proved by the Diebold–Mariano test that the prediction accuracy of random forest regression is significantly better than that of LSTM, the prediction errors RMSE and MAPE of random forest regression are better than those of LSTM. The changes in the variables that determine the price of Bitcoin in each period are also obtained through random forest regression. From 2015 to 2018, three US stock market indexes, NASDAQ, DJI, and S&P500 and oil price, and ETH price have impact on Bitcoin prices. Since 2018, the important variables have become ETH price and Japanese stock market index JP225. The relationship between accuracy and the number of periods of explanatory variables brought into the model shows that for predicting the price of Bitcoin for the next day, the model with only one lag of the explanatory variables has the best prediction accuracy.
Machine learning and deep learning algorithms produce very different results with different examples of their hyperparameters. Algorithm parameters require optimization because they aren't specific for all problems. In this paper Long Short-Term Memory (LSTM), eight different hyperparameters (go-backward, epoch, batch size, dropout, activation function, optimizer, learning rate and, number of layers) were used to examine to daily and hourly Bitcoin datasets. The effects of each parameter on the daily dataset on the results were evaluated and explained These parameters were examined with hparam properties of Tensorboard. As a result, it was seen that examining all combinations of parameters with hparam produced the best test Mean Square Error (MSE) values with hourly dataset 0.000043633 and daily dataset 0.00073843. Both datasets produced better results with the tanh activation function. Finally, when the results are interpreted, the daily dataset produces better results with a small learning rate and small dropout values, whereas the hourly dataset produces better results with a large learning rate and large dropout values.
In the last decade, the techniques of news aggregation and summarization have been increasingly gaining relevance for providing users on the web with condensed and unbiased information. Indeed, the recent development of successful machine learning algorithms, such as those based on the transformers architecture, have made it possible to create effective tools for capturing and elaborating news from the Internet. In this regard, this work proposes, for the first time in the literature to the best of the authors’ knowledge, a methodology for the application of such techniques in news related to cryptocurrencies and the blockchain, whose quick reading can be deemed as extremely useful to operators in the financial sector. Specifically, cutting-edge solutions in the field of natural language processing were employed to cluster news by topic and summarize the corresponding articles published by different newspapers. The results achieved on 22,282 news articles show the effectiveness of the proposed methodology in most of the cases, with 86.8% of the examined summaries being considered as coherent and 95.7% of the corresponding articles correctly aggregated. This methodology was implemented in a freely accessible web application.
B Sriman, Tamil Iniyal T, J Thasmiya, Taariq Ziyaadh J · 6 authors
India is rapidly moving towards the digitization of money in all aspects. Cryptocurrency has grown widely in India and around the world among investors for financial activities like buying, selling, and trading. According to the report submitted in 2021 by the United Nations Conference on Trade and Development, 7.3% of Indians owned cryptocurrency in 2021. In the past two years, i.e., 2020 and 2021, the value of global currencies have been falling due to the poor run of stock markets. So the investors found it very hard to cope with the economical issues. This in turn has led to a renewal of interest in digital currency. Our main target is to implement the efficient machine learning and deep learning- based models specifically Convolutional Neural Network (CNN), long short term memory(LSTM) and Gated Recurrent Units (GRU) to handle the price volatility of bitcoin and ethereum and to produce high accuracy.
Bitcoin was the first cryptocurrency introduced as a cryptographic proof-based electronic payment system in 2009. Till now approximately more than 10,000 digital coins are active in the crypto market. Cryptocurrency is a virtual digital asset that uses cryptography and blockchain technology for transaction verification and records maintenance. Its trading is gaining attention due to volatile behavior, decentralized nature, and liquidity in this digital asset. Trading this digital asset provides anonymity and security in transactions. Groundless fluctuations in its price contribute to making its trade risky. Market Prediction of the cryptocurrency is trending because it can reduce the trade loss risk. Data related to this market is vast and publicly available on the internet. It is nearly impossible to infer the market by simple data analysis. Statistical price prediction approaches are less effective due to the absence of seasonality in cryptocurrency market data. Therefore researchers proposed efficient price prediction techniques utilizing statistical, algorithmic, and neural network-based Machine Learning models. This paper provides a detailed literature survey related to the state-of-the-art Machine learning-based prediction methodologies for the market prediction of the digital asset from 2014 to 2022. This research will categorize, summarize, and review the existing research in cryptocurrency market prediction using Machine Learning classifiers. This paper will benefit researchers to be productive in the right direction in the future.
Machine learning has a wide range of applications to meet the complexity of data and various expectations for prediction types.In this study, a comprehensive review of various machine learning approaches for Bitcoin price prediction will be proposed.After examining previous research on cryptocurrency prediction using Long-Short Term Memory (LSTM), Multi-layer Perceptions (MLP), and Support Vector Machine (SVM), with the focus on LSTM, it can be found that LSTM is a widely employed method in Bitcoin price prediction because of its advantages in incorporating both long-term and short-term dependencies.This paper reviews a series of research papers by comparing the differences between the methods they implemented, to a limited extent, based on their predictive power, replicability, and model limitations.Furthermore, some potential improvements and explored innovations for future studies also be discussed.
Bu çalışmanın amacı, kripto para birimleri arasında piyasa değeri en yüksek olan Bitcoin ile diğer piyasa değeri en yüksek olan kripto para birimleri arasındaki fiyat hareketliliğinin incelenmesi ve birbirleriyle olan etkileşimin yönünün tespit edilmesidir. Ancak çalışmada piyasa değeri en yüksek olan 20 kripto para biriminden sadece 8 tanesinin verisine ulaşılabilmiştir. Verisine ulaşılabilen Bitcoin, Ethereum, Tether, BNB, XRP, Dogecoin, Tron, Ethereum Classic kripto paralarına ait 12.09.2017-01.09.2022 dönemini kapsayan 1815 adet veri Toda-Yamamoto nedensellik testi kullanılarak analiz edilmiştir. Yapılan analizler sonucunda, Bitcoin-Dogecoin, Bitcoin-Ethereum Classic, Bitcoin-Ethereum, Bitcoin -BNB, Bitcoin-Ripple, Bitcoin-Tron değişkenlerinin arasında çift yönlü bir nedensellik ilişki olduğu ancak Tether’den Bitcoin’e doğru tek yönlü bir nedensellik ilişkisi olduğu anlaşılmaktadır.
The financial software has expanded to include cryptocurrencies, which are seeing rapid adoption and are being positively received by critics. Mining is an essential part of these systems, which use a distributed ledger to store data in a trustworthy manner. The decentralized ledger, known as the blockchain is updated with information on prior transactions when mining is performed. Users are allowed to arrive at a reliable and robust agreement for each transaction. Mining can result in the generation of new wealth in the form of monetary assets, such as currency. Because cryptocurrencies were conceived from the outset to operate as decentralized, peer-to-peer networks, there is no centralized authority that can supervise the monetary transactions that take place using these currencies. Miners are accountable for ensuring that the transactions they are processing are legitimate. For crypto currencies, mining algorithms that are both dependable and strong are a fundamental must. This paper provides a comprehensive summary of crypto coins, specifically Bitcoin, Ethereum, and Litecoin, and an analysis and critique of the previous research on crypto currency trading that has been published. This paper presents a classification system that could be applied to both wellestablished standards and newly developed concepts.
Jan 1, 2023·Proceedings of the 4th Management Science Informatization and Economic Innovation Development Conference, MSIEID 2022, December 9-11, 2022, Chongqing, China
Quantitative portfolio of gold and bitcoin investment can be determined by synthesized quantitative model, with the help of various quantifying indicators. Previous prediction models labor to disperse risk of the investment portfolio as well as maximize the return. To handle it, this essay digs into