Abstract We study the statistical properties of the Bitcoin return series and provide a thorough forecasting exercise. Also, we calibrate stateâofâtheâart machine learning techniques and compare the results with econometric time series models. The empirical assessment provides evidence that the application of machine learning techniques outperforms econometric benchmarks in terms of forecasting precision for both inâ and outâofâsample forecasts. We find that both deep learning architectures as well as complex layers, such as LSTM, do not increase the precision of daily forecasts. Specifically, a simple recurrent neural network describes a sensible choice for forecasting daily return series.
S. Saraswathi, J S Sridhala, A. Elavazhagan, Jasbir Singh Sabharwal ¡ 5 authors
This research proposes an ensemble approach for Bitcoin price prediction, leveraging historical price data and sentiment analysis. The proposed ensemble approach combines the model with Gated Recurrent Unit (GRU) and Bidirectional Long Short-Term Memory (BiLSTM) to further improve the accuracy in prediction by considering dynamics in the market. The model also addresses the problem of generalization and overfitting, adaption to the changing, dynamic nature of the market. Historical price data and sentiment scores from the preprocessing of the text are combined to the ensemble framework. These data are then fed into GRU and BiLSTM models for training, as the data contain not only complex temporal patterns but also sentiment-driven trends. The ensemble strategy could be beneficial for the strengths of the models and for improving the performances of the predictors. Most importantly, features are engineered in terms of technical indicators, lagged variables, and external factors impacting the price of Bitcoin. Sentiment analysis with the news and on social media complements insight into market sentiment, which adds value to the prediction power of the model.
These days, there is a lot of demand for cryptocurrencies, and investors are essentially investing in them. The fact that there are already over 6,000 cryptocurrencies in use worldwide because of this, investors with regular incomes put money into promising cryptocurrencies that have low market values. Accurate pricing forecasting is necessary to build profitable trading strategies because of the unique characteristics and volatility of cryptocurrencies. For consistent forecasting accuracy in an unknown price range, a variation point detection technique is employed. Due to its bidirectional nature, a Bi-LSTM appropriate for recording long-term dependencies in data that is sequential. Accurate forecasting in the cryptocurrency space depends on identifying these connections, since values are subject to change over time due to a variety of causes. In this work, we employ four deep learning-based models that are LSTM, FB-Prophet, LSTM-GRU and Bidirectional-LSTM(Bi-LSTM) and these four models are compared with Silverkite. Silverkite is the main algorithm of the Python library Graykite by LinkedIn. Using historical bitcoin data from 2012 to 2021, we utilized to analyse the modelsâ mean absolute error (MAE) and root mean square error (RMSE). The Bi-LSTM model performs better than others, with a mean absolute error (MAE) of 0.633 and a root mean square error (RMSE) of 0.815. The conclusion has significant ramifications for bitcoin investors and industry experts.
Abstract Cryptocurrencies have garnered significant attention recently due to widespread investments. Additionally, researchers have increasingly turned to social media, particularly in the context of financial markets, to harness its predictive capabilities. Investors rely on platforms like Twitter to analyze investments and detect trends, which can directly impact the future price movements of Bitcoin. Understanding and analyzing Twitter sentiments can potentially provide insights into future Bitcoin price movements and can shed light on how investor sentiment affects cryptocurrency markets. In this study, we explore the correlation between Twitter activity and Bitcoin prices by examining tweets related to Bitcoin price sentiments. Our proposed model consists of two distinct networks. The first network exclusively utilizes historical price data, which is further decomposed into various components using the Empirical Mode Decomposition method. This decomposition helps mitigate the impact of irregular fluctuations on Bitcoin price predictions. Each of these components is then separately processed by Long Short-Term Memory (LSTM) networks. The second network focuses on modeling user sentiments and emotions in conjunction with Bitcoin market data. User opinions are categorized into positive and negative classes and are integrated with historical data to predict the next-day price using LSTM networks. Finally, the outputs of each network are combined to form the ultimate prediction values. Experimental results demonstrate that Twitter sentiment can effectively helps us predict Bitcoin price trends. Furthermore, to validate our proposed model, we compared it with several state-of-the-art methods. The results indicate that our approach outperforms these existing models in terms of accuracy.
The rapid growth and significant fluctuations of the cryptocurrency market have increasingly attracted investors to add digital currencies in their portfolios. Compared to traditional financial markets, the cryptocurrency market exhibits more pronounced characteristics of behavioral finance. Investors demonstrate irrational behavior in trading processes, exhibiting clear cognitive biases, such as the endowment effect and the ostrich effect. This paper initially undertakes an analytical dissection and synthesis of various archetypal irrational behaviors, then we selected technical indicators that reflect these irrational behaviors. After the process of feature engineering, the study employs TCN-MLP model to predict the thirty-minute returns of ETH. This paper presents a comprehensive cryptocurrency returns prediction process, addressing the weakness of loosely connected theory in previous research.
Unlike the traditional stock markets, the 24/7 nature of the cryptocurrency market poses unique challenges and opportunities, particularly in asset trading and management. These dynamic market conditions have accelerated the development of sophisticated trading strategies, increasingly leveraging the power of Artificial Intelligence (AI). Among these, AI-driven trading bots have become a prominent tool, offering enhanced decision-making capabilities over conventional methods. This paper proposes the application of the Advantage Actor-Critic (A2C) model, a reinforcement learning technique ideally suited for the unpredictable nature of the cryptocurrency market. Our research aims to optimize asset allocation within a diverse portfolio, including both high-volatility cryptocurrencies and the more stable US Dollar. The proposed A2C model strategically leverages current and predicted price data of cryptocurrencies with current asset allocation to make new asset allocation decisions. Our experiments demonstrate the A2C modelâs efficacy in managing asset allocations under varying market conditions. We particularly focus on how the model responds to alterations in the loss penalty factor within its reward function, which enables a shift between aggressive and conservative investment strategies. The model effectively balances risk and return, showing promising potential in achieving stable asset growth in rising markets while mitigating losses during market downturns.
Despite their recent inception, cryptocurrencies have become globally recognized for their dispersal, diversity, and high market capitalization. This volatility developed into a challenge for investors looking to predict price movements. Thus, it has become an attractive investment opportunity. To increase prediction accuracy, researchers integrate machine learning algorithms with technical indicators. In this review, a systematic comparison has been employed to identify efficient algorithms, and researchers have employed statistical measures to make short- and long-term forecasts of decentralized money prices. Moreover, the paper highlights the results of researchers based on machine learning and deep learning methodologies on multiple types of cryptocurrencies like Bitcoin, Ethereum, Monero, etc. Lastly, the work emphasizes the limitations, gaps, and challenges facing researchers to take advantage of existing literature for future works.
Manoel Fernando Alonso Gadi, MiguelâĂngel Sicilia
Abstract The objective of this paper is to describe Cryptocurrency Linguo (CryptoLin), a novel corpus containing 2683 cryptocurrency-related news articles covering more than a three-year period. CryptoLin was human-annotated with discrete values representing negative, neutral, and positive news respectively. Eighty-three people participated in the annotation process; each news title was randomly assigned and blindly annotated by three human annotators, one in each different cohort, followed by a consensus mechanism using simple voting. The selection of the annotators was intentionally made using three cohorts with students from a very diverse set of nationalities and educational backgrounds to minimize bias as much as possible. In case one of the annotators was in total disagreement with the other two (e.g., one negative vs two positive or one positive vs two negative), we considered this minority report and defaulted the labeling to neutral. Fleissâs Kappa, Krippendorffâs Alpha, and Gwetâs AC1 inter-rater reliability coefficients demonstrate CryptoLinâs acceptable quality of inter-annotator agreement. The dataset also includes a text span with the three manual label annotations for further auditing of the annotation mechanism. To further assess the quality of the labeling and the usefulness of CryptoLin dataset, it incorporates four pretrained Sentiment Analysis models: Vader, Textblob, Flair, and FinBERT. Vader and FinBERT demonstrate reasonable performance in the CryptoLin dataset, indicating that the data was not annotated randomly and is therefore useful for further research1. FinBERT (negative) presents the best performance, indicating an advantage of being trained with financial news. Both the CryptoLin dataset and the Jupyter Notebook with the analysis, for reproducibility, are available at the projectâs Github. Overall, CryptoLin aims to complement the current knowledge by providing a novel and publicly available Gadi and Ăngel Sicilia (Cryptolin dataset and python jupyter notebooks reproducibility codes, 2022) cryptocurrency sentiment corpus and fostering research on the topic of cryptocurrency sentiment analysis and potential applications in behavioral science. This can be useful for businesses and policymakers who want to understand how cryptocurrencies are being used and how they might be regulated. Finally, the rules for selecting and assigning annotators make CryptoLin unique and interesting for new research in annotator selection, assignment, and biases.
The current investigation delves into the valuation trends of prominent Non-Fungible Token (NFT) collections, entities at the forefront of the digital economy that are redefining concepts of asset ownership and artistic appreciation. The academic import of this study is anchored in the emergent nature of NFTs and their paradigmatic shift from traditional economic models. This research aims to decipher the market value fluctuations of top NFT collections through a comprehensive time-series analysis, the application of the ARIMA model. Methodologically, the study adheres to established time-series analytical procedures, with a focus on identifying and interpreting patterns and trends within the market data. The approach synthesizes a broad spectrum of economic and cultural variables that potentially exert influence over the NFT marketplace. Results gleaned from the analysis yield insights into the price movements of NFTs, offering a juxtaposition of predicted and actual market data to deepen the understanding of this unique market sector. The implications of this research extend beyond academic interest, offering a vital resource for stakeholders within the digital economy. The discerned patterns and dynamics of NFTs, as revealed through the study, contribute to a granular understanding of digital assets, providing a scaffold for future research endeavors and practical guidance for participants in the digital art and ownership space.
Nehal N. AlMadany, Omar Hujran, Ghazi AlâNaymat, Aktham Maghyereh
The emergence of cryptocurrencies has generated enthusiasm and concern in the modern global economy. However, their high volatility, erratic price fluctuations, and tendency to exhibit price bubbles have made investors cautious about investing in them. Consequently, it is essential to develop methods and models to forecast cryptocurrency returns to benefit investors, traders, and the scientific community. Despite the considerable volume of research on Bitcoin price forecasting, other cryptocurrencies have received little attention in academic literature. Additionally, the current body of literature on predicting cryptocurrency prices or returns emphasizes the use of in-sample methodologies. However, this method is susceptible to overfitting. To address these gaps in the literature, this study employs autoregressive moving average (ARMA), generalized autoregressive conditional heteroskedasticity (GARCH), exponential generalized autoregressive conditional heteroskedasticity (EGARCH), and long short-term memory (LSTM) deep learning neural networks to forecast returns for the ten most actively traded digital currencies: Bitcoin, Ethereum, Ripple, Chainlink, Litecoin, Cardano, Ethereum Classic, Bitcoin Cash, Tether, and Binance Coin. To assess the accuracy of the two models, this study utilizes an out-of-sample method with data gathered sequentially from November 9, 2017, to September 18, 2022. The results indicate that all models exhibit high accuracy, as evidenced by their low root mean square error (RMSE), mean absolute error (MAE), and mean squared error (MSE) values. Meanwhile, the hybrid EGARCH-LSTM or GARCH-LSTM models demonstrate slightly better accuracy compared with the other models. The findings are valuable for investors, traders, and researchers involved in cryptocurrency forecasting.
The system proposed in this paper aims to predict cryptocurrency prices using Bi-Directional Long Short- Term Memory (LSTM), leveraging historical data obtained from Yahoo Finance and CoinGecko APIs. The goal is to assess LSTM models effectiveness in forecasting cryptocurrency prices and offer an interactive interface for users to visualize historical and forecasted prices. Several research works have been conducted on the prediction of cryptocurrency prices through various Deep Learning (DL) based algorithms. This project comprises two main approaches : one involves data analysis, LSTM modeling, and change point detection using Yahoo Finance data, while the other focuses on LSTM model training and price prediction using CoinGecko API data. The paper suggests that the prediction models it presents are useful for traders, investors, [6] and finance academics and are close to accurate at predicting the values of cryptocurrencies. Future research will examine more advanced deep learning architectures, primarily Transformer-based models like the GPT series, to improve pattern detection in bitcoin data. Integrating other data sources, such as sentiment analysis or blockchain measurements, may increase the accuracy of forecasting. With further research into cutting-edge techniques, cryptocurrency forecasting will get better and provide stakeholders with more information to help them make informed decisions. Keywordsâ Cryptocurrency ; forecasting ; Bi-Directional LSTM Model ; Time-series forecasting ; Machine learning
Financial market prediction has profound impacts on trading, investing, and risk management strategies. However, accurate forecasting of asset prices like stocks, cryptocurrencies, forex, and commodities remains challenging due to the complexity, noise, and non-stationarity in financial time series data. This project leverages machine learning to develop data-driven predictive models that can capture intricate patterns and provide actionable insights. Three distinct algorithms are explored - Moving Average Convergence Divergence (MACD) for next-day price forecasting, linear regression for next-day forecasting, and Long Short-Term Memory (LSTM) recurrent neural networks for predicting prices over a 1-week (7-day) horizon. Extensive historical data from major stock indices like S&P 500, NIFTY 50, top cryptocurrencies, forex currency pairs, and commodities are utilized for training and evaluation. The models incorporate both technical indicators derived from price/volume data as well as fundamental factors and news sentiment obtained from sources like Alpha Vantage API. A rigorous methodology involving data preprocessing, feature engineering, model training, hyperparameter optimization, and backtesting on unseen data is employed to ensure the models' robustness and generalization capabilities. Appropriate error metrics like mean squared error and directional accuracy are used for quantitative performance assessment. Additionally, interpretability techniques are applied to the LSTM models to uncover non-linear patterns and understand the key drivers influencing the forecasts. The overarching goal is to develop accurate predictive systems that financial institutions, quantitative funds, and individual investors can leverage for applications like algorithmic trading, portfolio optimization, and data-driven investment decision support across different asset classes and market conditions
Abstract Nowadays, the prediction of cryptocurrency side effects on the critical aspects of the exchange rates in intelligent business is one of the main challenges in the financial market. Cryptocurrency is defined as a set of digital information concerning internal financial protocols of digital marketing, such as blockchain, which operates according to a decentralized architecture. On the other hand, fraud activities in Ethereum transfer and management of cryptocurrency now increase and affect safe transactional processes. This article presents a new machineâlearning approach to Ethereum fraud Detection based on Bayesian Optimizable Ensemble Bagged Trees (BOEBT) algorithm. Moreover, the main goal of this study is to derive the accuracy of the cryptocurrency prediction model using different machineâlearning algorithms and compare their evaluation parameters together. The performance of the proposed prediction model using the machine learning algorithms was evaluated by the MATLAB tool. The experimental results show that the proposed BOEBT algorithm merits achieving 99.21% accuracy and 99.14% F1âScore to other machine learning algorithms for cryptocurrency fraud prediction.
In the volatile cryptocurrency market, accurately forecasting Bitcoin prices is crucial yet challenging, carrying significant economic implications. This paper presents a novel hybrid model that merges the predictive capabilities of Long Short-Term Memory (LSTM) networks with the computational efficiency of Gated Recurrent Units (GRU). This integration is designed to simultaneously capture long-term dependencies and short-term fluctuations inherent in Bitcoin price dynamics, thus providing a comprehensive analysis framework. The model utilizes a meticulous architecture starting with an input layer that normalizes data to address price variability, followed by LSTM layers that interpret long-term trends, and GRU layers that refine insights based on short-term variations. Evaluated using a dataset divided into training, validation, and testing phases and optimized with the Adam algorithm, the modelâs performance surpasses traditional forecasting methods and standalone neural networks. Metrics such as RMSE, MAE, and R2confirm its superior predictive accuracy, with significant improvements over benchmarks like ARIMA, standalone LSTM, and GRU models. This breakthrough highlights the Hybrid LSTM-GRU modelâs potential as a transformative tool for investors and analysts navigating the complexities of the cryptocurrency market.
As the non-fungible token (NFT) market flourishes, price prediction emerges as a pivotal direction for investors gaining valuable insight to maximize returns. However, existing works suffer from a lack of practical definitions and standardized evaluations, limiting their practical application. Moreover, the influence of users' multi-behaviour transactions that are publicly accessible on NFT price is still not explored and exhibits challenges. In this paper, we address these gaps by presenting a practical and hierarchical problem definition. This approach unifies both collection-level and token-level task and evaluation methods, which cater to varied practical requirements of investors. To further understand the impact of user behaviours on the variation of NFT price, we propose a general wallet profiling framework and develop a COmmunity enhanced Multi-bEhavior Transaction graph model, named COMET. COMET profiles wallets with a comprehensive view and considers the impact of diverse relations and interactions within the NFT ecosystem on NFT price variations, thereby improving prediction performance. Extensive experiments conducted in our deployed system demonstrate the superiority of COMET, underscoring its potential in the insight toolkit for NFT investors.
Kia Jahanbin, Mohammad Ali Zare Chahooki, Mahdi YazdianâDehkordi, Fatemeh Rahmanian
OBJECTIVES: Due to the limitations of Twitter, the expansion of Telegram channels, and the Telegram API's easy use, Telegram comments have become prevalent. Telegram is one of the most popular social networks, unlike Twitter, which has no restrictions on sending messages, and experts can share their opinions and media. Some of these channels, managed by influencers of large companies, are very influential in the behavior of the market on various stocks, including cryptocurrencies. In this research, the opinion collection of 10 famous Telegram channels regarding the analysis of cryptocurrencies has been extracted. The sentiments of these opinions have been analyzed using the HDRB model. HDRB is a hybrid model of RoBERTa deep neural network, BiGRU, and attention layer used for sentiment analysis (SA). Analyzing the sentiments of these opinions is very important for understanding the future behavior of the market and managing the stock portfolio. The opinions of this dataset, published by experts in the field of cryptocurrencies, are precious, unlike the opinions that are extracted only by using the hashtag of the names of cryptocurrencies. On the other hand, the dataset related to cryptocurrencies, which has the opinions of experts and the polarity of their feelings, is very rare. DATA DESCRIPTION: The dataset of this research is the sentiments of more than ten popular Telegram channels regarding a wide range of cryptocurrencies. These comments were collected through the Telegram API from December 2023 to March 2024. This data set contains an Excel file containing the text of the comments, the date of comment creation, the number of views, the compound score, the sentiment score, and the type of sentiment polarity. These opinions cover influencer analysis on a wide range of cryptocurrencies. Also, two Word files, one containing the description of the dataset columns and the other Python code for extracting comments from Telegram channels, are included in this dataset.
Joy Dip Das, Ruppa K. Thulasiram, Christopher J. Henry, A. Thavaneswaran
This work addresses the intricate task of predicting the prices of diverse financial assets, including stocks, indices, and cryptocurrencies, each exhibiting distinct characteristics and behaviors under varied market conditions. To tackle the challenge effectively, novel encoderâdecoder architectures, AE-LSTM and AE-GRU, integrating the encoderâdecoder principle with LSTM and GRU, are designed. The experimentation involves multiple activation functions and hyperparameter tuning. With extensive experimentation and enhancements applied to AE-LSTM, the proposed AE-GRU architecture still demonstrates significant superiority in forecasting the annual prices of volatile financial assets from the multiple sectors mentioned above. Thus, the novel AE-GRU architecture emerges as a superior choice for price prediction across diverse sectors and fluctuating volatile market scenarios by extracting important non-linear features of financial data and retaining the long-term context from past observations.
Raed Alsini, Qasem Abu AlâHaija, Abdulaziz A. Alsulami, Badraddin Alturki ¡ 8 authors
Introduction: The cryptocurrency market is captivating the attention of both retail and institutional investors. While this highly volatile market offers investors substantial profit opportunities, it also entails risks due to its sensitivity to speculative news and the erratic behavior of major investors, both of which can provoke unexpected price fluctuations. Methods: In this study, we contend that extreme and sudden price changes and atypical patterns might compromise the performance of technical signals utilized as the basis for feature extraction in a machine learning-based trading system by either augmenting or diminishing the model's generalization capability. To address this issue, this research uses a bagged tree (BT) model to forecast the buy signal for the cryptocurrency market. To achieve this, traders must acquire knowledge about the cryptocurrency market and modify their strategies accordingly. Results and discussion: To make an informed decision, we depended on the most prevalently utilized oscillators, namely, the buy signal in the cryptocurrency market, comprising the Relative Strength Index (RSI), Bollinger Bands (BB), and the Moving Average Convergence/Divergence (MACD) indicator. Also, the research evaluates how accurately a model can predict the performance of different cryptocurrencies such as Bitcoin (BTC), Ethereum (ETH), Cardano (ADA), and Binance Coin (BNB). Furthermore, the efficacy of the most popular machine learning model in precisely forecasting outcomes within the cryptocurrency market is examined. Notably, predicting buy signal values using a BT model provides promising results.
This electronic document describes the involvement of finance agencies and various types of corporate industries through âsyndicated loansâ that simplify financial transactions for both agencies and corporations. This will be a revolutionary step towards the âfintechâ sector as it is all about decentralization and how it changes the whole game during the lending process. This paper describes a proposed process of lending, monitoring, evaluation, and development in the lending process between borrowers (enterprises) and lenders using Large Language Model. After highlighting the difficulties in implementing this proposal, the paper also discusses its impact on the transition process and outlines what has already been done. The proposal increases the transparency of the entire process among stakeholders, simplifies manual work, reduces operational costs, and provides data to support the analysis and summarization of benefits derived from the lending process. A secure system offers the ease of use and dexterity of the technology that makes the agenda highly reliable in all aspects of test cases. The study includes the experimentation of proposed system that how it reduces the failure rate of startups by 70% by providing financial aids to startups and suggesting a better business model if needed.
Purpose Candlestick charts are a key tool for the technical analysis of cryptocurrency price fluctuations. It is essential to examine trends in the time series of a financial asset when completing an analysis. To accurately examine its potential future performance, it must also consider how it has changed and been active during the period. The researchers created cryptocurrency trading algorithms in this study based on the traditional candlestick pattern. Design/methodology/approach The data includes information on Bitcoin prices from early 2012 until 2021. Only the engulfing Candlestick model was able to anticipate changes in the price movements of Bitcoin. The traditional Harami model does not work with Bitcoin trading platforms because it has yet to generate profitable business results. An inverted Harami is a successful cryptocurrency trading method. Findings The inverted Harami approach accounts for 6.98 profit factor (PrF) and 74â50% of profitable (Pr) transactions, which favors a particularly long position. Additionally, the study discovered that almost all analyzed candlestick patterns forecast longer trends greater than shorter trends. Research limitations/implications To statistically study its future potential return, examining how it has changed and been active over the years is necessary. Such valuations are the basis for trading strategies that could help traders and investors in the cryptocurrency market. Without sacrificing clarity or ease of application, the proposed approach has increased performance by up to 32.5% of mean absolute error (MAE). Originality/value This study is novel in that it used multilayer autoregressive neural network (MARN) models with crypto-net (CNM) in machine learning to analyze a time series of financial cryptocurrencies. Here, the primary study deals with time trends extracted through a neural network model. Then, the developed model was tested using Bitcoin and Ethereum. Finally, CNM validity was tested through linear regression.