Radhakrishna Dodmane, K. R. Raghunandan, Krishnaraj Rao N S, Bhavya Kallapu · 7 authors
The advancements in communication speeds have enabled the centralized financial market to be faster and more complex than ever. The speed of the order execution has become exponentially faster when compared to the early days of electronic markets. Though the transaction speed has increased, the underlying architecture or models behind the markets have remained the same. These models come with their own disadvantages. The disadvantages are usually faced by non-institutional or small traders. The bigger players, such as financial institutions, have an advantage over smaller players because of factors such as information asymmetry and access to better infrastructure, which give them an advantage in terms of the speed of execution. This makes the centralized stock market an uneven playing field. This paper discusses the limitations of centralized financial markets, particularly the disadvantage faced by non-institutional or small traders due to information asymmetry and better infrastructure access by financial institutions. The authors propose the usage of blockchain technology and the data highway protocol to create a decentralized stock exchange that can potentially eliminate these disadvantages. The data highway protocol is used to generate new blocks with a flexible finality condition that allows for the consensus mechanism to configure security thresholds more freely. The proposed framework is compared with existing frameworks to confirm its effectiveness and identify areas that require improvement. The evaluation of the proposed approach showed that the improved highway protocol boosted the transaction rate compared to the other two mechanisms (PoS and PoW). Specifically, the transaction rate of the proposed model was found to be 2.2 times higher than that of PoS and 12 times higher than that of the PoW consensus model.
M. Karthik, Vethath Suryakumar. R, S Syedjaffar., Dharneesh. M.E
Abstract: Our project aim is to predict the future price of the bitcoin using machine learning algorithms. In the modern world cryptocurrency is become more trendy and reaches the youngsters to invest in the stock market and to generate some profitable trades. To invest their money in cryptocurrency we are just helping the investors like peoples and also involving the organization to invest in the bitcoin and to make good profitable trades. Initially it utilizes the historical data to predict the future price of the bitcoin. It involves considering factors such as market sentiment, news and events ,technical analysis, and global economic trends. Different machine learning algorithms are applied on the a data and the accuracy is compared to see which algorithm performed better. It includes the performance metrics like precision, recall scores are also taken into consideration for evaluating the model In cryptocurrency market it contains 'n' number of coins .Among those coins we can take any coin to predict the future price which will able to help the investors who are all investing their money in cryptocurrency market.
Abstract Anomalies, which are incompatible with the efficient market hypothesis and mean a deviation from normality, have attracted the attention of both financial investors and researchers. A salient research topic is the existence of anomalies in cryptocurrencies, which have a different financial structure from that of traditional financial markets. This study expands the literature by focusing on artificial neural networks to compare different currencies of the cryptocurrency market, which is hard to predict. It aims to investigate the existence of the day-of-the-week anomaly in cryptocurrencies with feedforward artificial neural networks as an alternative to traditional methods. An artificial neural network is an effective approach that can model the nonlinear and complex behavior of cryptocurrencies. On October 6, 2021, Bitcoin (BTC), Ethereum (ETH), and Cardano (ADA), which are the top three cryptocurrencies in terms of market value, were selected for this study. The data for the analysis, consisting of the daily closing prices for BTC, ETH, and ADA, were obtained from the Coinmarket.com website from January 1, 2018 to May 31, 2022. The effectiveness of the established models was tested with mean squared error, root mean squared error, mean absolute error, and Theil’s U1, and $${R}_{OOS}^{2}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msubsup> <mml:mi>R</mml:mi> <mml:mrow> <mml:mi>OOS</mml:mi> </mml:mrow> <mml:mn>2</mml:mn> </mml:msubsup> </mml:math> was used for out-of-sample. The Diebold–Mariano test was used to statistically reveal the difference between the out-of-sample prediction accuracies of the models. When the models created with feedforward artificial neural networks are examined, the existence of the day-of-the-week anomaly is established for BTC, but no day-of-the-week anomaly for ETH and ADA was found.
Gas is the transaction-fee metering system of the Ethereum network. Users of the network are required to select a gas price for submission with their transaction, creating a risk of overpaying or delayed/unprocessed transactions involved in this selection. In this work, we investigate data in the aftermath of the London Hard Fork and shed insight into the transaction dynamics of the network after this major fork. As such, this paper provides an update on work previous to 2019 on the link between EthUSD/BitUSD and gas price. For forecasting, we compare a novel combination of machine learning methods such as Direct-Recursive Hybrid LSTM, CNN-LSTM, and Attention-LSTM. These are combined with wavelet threshold denoising and matrix profile data processing toward the forecasting of block minimum gas price, on a 5-min timescale, over multiple lookaheads. As the first application of the matrix profile being applied to gas price data and forecasting that we are aware of, this study demonstrates that matrix profile data can enhance attention-based models; however, given the hardware constraints, hybrid models outperformed attention and CNN-LSTM models. The wavelet coherence of inputs demonstrates correlation in multiple variables on a 1-day timescale, which is a deviation of base free from gas price. A Direct-Recursive Hybrid LSTM strategy is found to outperform other models, with an average RMSE of 26.08 and R2 of 0.54 over a 50-min lookahead window compared to an RMSE of 26.78 and R2 of 0.452 in the best-performing attention model. Hybrid models are shown to have favorable performance up to a 20-min lookahead with performance being comparable to attention models when forecasting 25–50-min ahead. Forecasts over a range of lookaheads allow users to make an informed decision on gas price selection and the optimal window to submit their transaction in without fear of their transaction being rejected. This, in turn, gives more detailed insight into gas price dynamics than existing recommenders, oracles and forecasting approaches, which provide simple heuristics or limited lookahead horizons.
This paper aims to analyze cryptocurrency volatility by examining the effect of Gold, Dollar Index, and Composite Stock Price Index (IHSG) as independent variables and on Bitcoin and Ethereum as dependent variables. The cryptocurrency objects in this study are Bitcoin and Ethereum, which have the largest market capitalization. The data in this study used the period January 1, 2018, to December 31, 2021. This study used GARCH analysis. This study's results indicate that Bitcoin's volatility is influenced by the price of Bitcoin itself, gold, and the stock exchange index, and Ethereum and the stock exchange index influence Ethereum. This shows that the cryptocurrency market is inefficient as the prices are also affected by past prices.
Cryptocurrencies are digital currencies that operate on the blockchain, which is the technology that offers security and decentralization. The principal characteristic of cryptocurrencies is that they are not generally issued by a central authority. Many factors can influence the volatility of prices. This paper enables to drive insights into the behavior of markets through the application of sentiment analysis of Tweets, Google news and machine learning techniques for the challenging task of cryptocurrency price prediction. Most of the studies have focused exclusively on the sentiment analysis of tweets. In this work, we propose the use of common machine learning tools and available Google News data for predicting the price of crypto. We present the results of the Long Short-Term Memory (LSTM) model using Tweets and Google News data.
Mr. R. Arunachalam, Myana Santhoshini, R. Tamil Prabha, R. Tamil Prabha
In this paper, we tried to estimate the Bitcoin price precisely taking into consideration various parameters that affect the Bitcoin value. In our work, we pointed to understand and identify daily changes in the Bitcoin market while obtaining insight into most appropriate features surrounding Bitcoin price. We will predict the daily price change with highest possible accuracy. The market capitalization of publicly traded cryptocurrencies is currently above $230 billion. Bitcoin, the most valuable cryptocurrency, serves primarily as a digital store of value, and its price predictability has been well-studied. For the first phase of our investigation, we aim to understand and identify daily trends in the Bitcoin market while gaining insight into optimal features surrounding Bitcoin price. Our data set consists of various features relating to the Bitcoin price and payment network over the course of five years, recorded daily. For the second phase of our investigation, using the available information, we will predict the sign of the daily price change with highest possible accuracy with deep learning algorithm such as long short term memory for greater accuracy. Compared with benchmark results for daily price prediction, we achieve a better performance, with the highest accuracies of the statistical methods and deep learning algorithms. Deep Learning models includes Long Short-Term Memory in RNN for Bitcoin price prediction are superior to statistical methods
Aims: This article investigates recent advancements in machine learning and blockchain technology for cryptocurrency price prediction. The study presents a ML system using various techniques applied to six different datasets. The findings highlight that simpler models can outperform complex ones in predicting cryptocurrency prices. Methods: The methods used in this study include applying diverse ML techniques such as LSTM, CNN, SVM, KNN, XGBoost, Astro ML, LASSO, RIDGE, linear regression, DT, and GP on six cryptocurrency datasets to predict prices. Results: The research evaluated various machine learning techniques for predicting cryptocurrency prices and reported the following RMSE values: Bitcoin prediction using Nadaraya-Watson kernel regression yielded an RMSE of 0.17, while Dogecoin prediction with linear regression resulted in an RMSE of 0.032. Ethereum price prediction using Gaussian regression achieved an RMSE of 0.02. For USD Coin, a combination of XGBoost, Gaussian regression, and Ridge techniques led to an RMSE of 0.014. Binance Coin price prediction using Gaussian regression had an RMSE of 0.032, and finally, Cardano Coin prediction employing LSTM reached an RMSE of 0.059. Conclusion: This study demonstrated the effectiveness of various machine learning techniques in predicting cryptocurrency prices. It revealed that simpler models can outperform complex ones in certain cases. The research contributes valuable insights to the field and can guide future work in cryptocurrency price prediction. The proposed model achieved promising results as evaluated by the RMSE metric.
Teknolojinin gelişmesiyle birlikte kripto para borsaları insanların daha fazla gelir elde etmek amacıyla kullandığı borsalardan biri olmuştur. Borsalarda alım-satım işlemleri yapılırken teknik ve temel analiz yöntemleri kullanılmaktadır. Teknik analiz, geçmiş verilerden yola çıkarak gelecekteki fiyat hareketlerini tahmin etme işlemidir. Teknik analiz yapılırken çok büyük verilerle karşılaşılınca verilerin analizi zorlaşmakta ve teknik analiz sonucu elde edilecek verilerin hatalı olma ihtimali artmaktadır. Bu durum sonucunda büyük verileri doğru analiz edemeyen yatırımcıların büyük zararlara uğrama ihtimali artmaktadır. Kripto para tahmini hem yatırımcılara doğru karar almak için hem de bilimsel alanda uygulamalara açık olduğu için değerlidir. Bu sebeple bu çalışmada, kripto para hareketliliği en yüksek olan kripto paralar arasından 3 adet kripto para seçilerek fiyat tahmini çalışması yapılmıştır. Seçilen kripto paralar; Bitcoin, Ethereum ve Cardano’dur. Verilerin büyük olması sebebiyle ve karar etkenlerinin analizi açısından Yapay Sinir Ağları ve Regresyon Analizi yöntemleri ile bu kripto paraların açılış, kapanış, gün içindeki en küçük ve en büyük değerleri kullanılarak bir sonraki günün kapanış değeri tahmin edilmiştir. Sonrasında tahmini değerlerle gerçek değerler arasında karşılaştırma yapılmıştır. Çalışma sonucunda Yapay Sinir Ağları ile yapılan tahmin çalışmasının Regresyon Analizi ile yapılan tahmin çalışmasından daha başarılı performans sergilediği gözlemlenmiştir.
The aim of this paper is to investigate the effect of a novel method called linear law-based feature space transformation (LLT) on the accuracy of intraday price movement prediction of cryptocurrencies. To do this, the 1-minute interval price data of Bitcoin, Ethereum, Binance Coin, and Ripple between 1 January 2019 and 22 October 2022 were collected from the Binance cryptocurrency exchange. Then, 14-hour nonoverlapping time windows were applied to sample the price data. The classification was based on the first 12 hours, and the two classes were determined based on whether the closing price rose or fell after the next 2 hours. These price data were first transformed with the LLT, then they were classified by traditional machine learning algorithms with 10-fold cross-validation. Based on the results, LLT greatly increased the accuracy for all cryptocurrencies, which emphasizes the potential of the LLT algorithm in predicting price movements.
The growing potential and high volatility of the cryptocurrency market attract a lot of interest from both businesses and investors. Even though the prices fluctuate, predicting with time serious models such as ARMA and ARIMA would still provide a useful reference for analyzing the market. Recent studies on machine learning methods including RNNs have made new progress in forecasting digital currencies. This study focuses on one of the traditional models ARMA to predict the time serious dataset from 2021-2022 of cryptocurrencies including Bitcoin, Ethereum and Ripple. To be specific, AIC and ADF tests are used to choose the optimal model and suitable dataset. According to the analysis, the ARMA model would be affected by the volatility of Bitcoin. However, the predictions are not precise enough but still a valuable reference for certain businesses and individual investors. More state-of-art machine learning models can be utilized in future study to enhance the performance. Overall, these results shed light on guiding further exploration of crypto currency price prediction.
Yapılan bu çalışmanın amacı, 1.1.2016-4.12.2022 dönemini kapsayan günlük veriler yardımıyla Bitcoin ile alakalı çıkan haberler ile hem Bitcoin fiyatı hem de getirisi arasındaki ilişkiyi zamanla değişen nedensellik analizi kapsamında incelemektir. Söz konusu ilişkinin varlığı, Hacker ve Hatemi-J (2006)’nin Boostrapt Temelli Toda-Yamamoto Nedensellik Testi ve zamanla değişen nedensellik analizi kullanılarak araştırılmıştır. Elde edilen nedensellik testi bulguları, Bitcoin ile ilgili çıkan haberler ile Bitcoin fiyatı arasında karşılıklı bir nedensellik ilişkisi olduğu yönündedir. Diğer taraftan, Bitcoin getirisi ile Bitcoin ile ilgili çıkan haberler arasındaki nedensellik bulguları incelendiğinde, Bitcoin ile alakalı çıkan haberlerden Bitcoin getirisine doğru nedensellik ilişkisinin söz konusu olmadığı, buna karşın Bitcoin getirilerinden Bitcoin ile alakalı çıkan haberlere doğru bir nedensellik olduğu söylenebilir. Ayrıca, söz konusu nedensellik ilişkilerinin zamanla nasıl bir seyir izlediğine bakıldığında özellikle Bitcoin fiyatlarının arttığı dönemlerde Bitcoin ile ilgili haber sayılarının arttığı görülmüştür. Bu çerçevede hem Bitcoin hem de altcoin piyasasına yatırım yapacak bireylerin, Bitcoin ve altcoin ile alakalı çıkmış olan haberleri dikkate alarak işlem yapmaları yatırımın sağlıklı olması adına önem teşkil etmektedir.
Vasileios Kochliaridis, Eleftherios Kouloumpris, Ioannis Vlahavas
Abstract Cryptocurrency markets experienced a significant increase in the popularity, which motivated many financial traders to seek high profits in cryptocurrency trading. The predominant tool that traders use to identify profitable opportunities is technical analysis. Some investors and researchers also combined technical analysis with machine learning, in order to forecast upcoming trends in the market. However, even with the use of these methods, developing successful trading strategies is still regarded as an extremely challenging task. Recently, deep reinforcement learning (DRL) algorithms demonstrated satisfying performance in solving complicated problems, including the formulation of profitable trading strategies. While some DRL techniques have been successful in increasing profit and loss (PNL) measures, these techniques are not much risk-aware and present difficulty in maximizing PNL and lowering trading risks simultaneously. This research proposes the combination of DRL approaches with rule-based safety mechanisms to both maximize PNL returns and minimize trading risk. First, a DRL agent is trained to maximize PNL returns, using a novel reward function. Then, during the exploitation phase, a rule-based mechanism is deployed to prevent uncertain actions from being executed. Finally, another novel safety mechanism is proposed, which considers the actions of a more conservatively trained agent, in order to identify high-risk trading periods and avoid trading. Our experiments on 5 popular cryptocurrencies show that the integration of these three methods achieves very promising results.
Nurazlina Abdul Rashid, Mohd Tahir Ismail, Noor Wahida Md Junus
Predicting cryptocurrency prices are difficult due to dynamic data. At the same time, the hidden market behavior of trend and seasonal components in the history data is also critical as it provides an idea of what the price pattern will be in the future. Hence, this research proposes to identify and model the hidden pattern behavior in terms of component time series instead of removing it via the linear structural time series (STS) model approach. This study focuses on the top five cryptocurrencies relying on the highest market capitalization. From the results obtained, the top five cryptocurrencies have a different trend model, either deterministic or stochastic, which relies on the behavior of data. The five cryptocurrencies also show the crypto winter event, where the trend is downward after six months every year. The linear STS is the best model for predicting three cryptocurrencies’ prices for nonstationary and volatility data behavior. It can also handle the hidden component behavior and is easy to interpret. Since the linear STS model can indirectly retain the information of data, it will assist investors and traders in accurately predicting cryptocurrency prices.
As the price of virtual currency fluctuates greatly, precise prediction and appropriate trading strategies can bring investors best returns. This paper predicted the price of Ethereum and Bitcoin in the light of autoregressive integrated moving average model (ARIMA) and get a R2 of 0.995 and 0.993 respectively, which indicates the model can yield reasonable predictions. Then their investment ratios are set to 0.88 and 1.12 respectively by analytic hierarchy process (AHP). Particle swarm optimization (PSO) is used to solve the daily revenue function formed by the predicted price and the current price. Finally, the paper compared the returns yielded by the PSO trading strategy optimized by AHP and the strategy without optimization. It can be concluded that the AHP has a possibility of 64.66 per cent to yield more returns when used.
The combination of Deep Learning and GARCH-type models has been proved to be superior to the single models in forecasting of volatility in various markets such as energy, main metals, and especially stock markets. To verify this hypothesis for cryptocurrencies market, we constructed various Deep Learning models based on Feed Forward Neural Networks (DFFNNs) and Long Short-Term Memory (LSTM) networks and evaluated their performance in forecasting the volatility of 27 cryptocurrencies. Then, different hybrid models were built in which the outputs of three GARCH-type models, namely GARCH, EGARCH, and APGARCH, with three different assumptions for the residuals’ distribution were fed into the DFFNN and LSTM networks. In other words, GARCH-type models were utilized as feature extractors and the deep learning models leveraged a sequence of extracted features as their inputs to produce the volatility of the next day. Our findings revealed that not only the deep learning models improve the forecasts of GARCH-type models with any distribution assumption, the forecasts of GARCH-type models as informative features can significantly increase the predictive power of the studied deep learning models; namely, the DFFNN and LSTM models.
The price of cryptocurrency is always volatile and is influenced by various factors like market returns, prices of stocks, gold, and correlation of prices of cryptocurrency. Modeling and forecasting the prices of cryptocurrencies and measuring the volatility with the GARCH specification (Engle, 1982) has become standard among researchers. Several applications and extensions of GARCH model is proposed by Bollerslev (1986). Later, an integrated GARCH model (Engle & Bollerslev, 1986) states that the persistence parameter is equal to one. A combination of short and long memory conditional models for the mean and the volatility to analyze crypto returns is done with the help of ARFIMA (Autoregressive Fractionally Integrated Moving Average) and FIGARCH (Fractionally Integrated Generalized Autoregressive Conditionally Heteroskedastic) Model. This paper intended to understand various mathematical models for volatility of crypto currencies and also to find research gaps in the existing literature. A comprehensive overview is the need of the study.
<p>We extend the Shariah-compliant digital assets and Islamic Fintech literature through exploring the time-frequency associations between the volatility index (VIX) and cryptocurrencies (both Islamic and traditional). Employing wavelet-based technique, we find that Islamic cryptocurrencies demonstrate low or no coherency with stock market volatility compared to traditional cryptocurrencies (except Tether) during the whole time and frequency bands, highlighting the hedging capabilities of Islamic cryptocurrencies. Tether also serves the same against VIX, as there is a low or favorable link between these variables. Finally, our findings would be prolific to digital currency traders and investors in designing the portfolio strategies.</p>
With growing interest toward investments in the cryptocurrency market, prediction of the volatility of the price increasingly becomes important. Given the popularity of social media activity to reflect market trends in recent years, sentiment analysis has been recognized as a great contributing factor to predict financial markets. Using a sample of Bitcoin and Ethereum trade data, this study intends to provide insights on the association between twitter activity about cryptocurrencies and fluctuations of their price. To this end, we implement regression analysis alongside Vector Autoregression method to examine to what extent sentiment-related measures are capable of explaining the volatility of the prices of cryptocurrencies and whether the mutual influence of sentiment and volatility improves the accuracy of the model. Results indicate that the accuracy of predictions vary across the two tested cryptocurrencies, and also two different lexicon approaches used to calculate sentiment scores.
Contemporarily, under the impacts of COVID-19 and regional conflicts with radicalness fiscal policy, the prices of cryptocurrency have been fluctuated dramatically. Among various types of cryptocurrency, Ethereum is one of the most volatility assets. In order to avoid risks as well as gain extra return in the crypto market, it is necessary to construct accurate prediction approach. In this paper, the Long Short-Term Memory algorithm will be used to predict the future price of Ethereum by learning Ethereum's past price direction data. price trend by learning Ethereum's past price trend data. Based on the analysis, the predicted values of the trained model fit well with the actual data, with the regression evaluation index R2 of 97.08% and MAPE of 6.89%. According to the results, it is feasible to predict the future price trend through the past price trend data. Nevertheless, it should be noted that the stochastic process in data training might lead to the instability of model performances. Hence, it is necessary to train the data several time to select the best models. Overall, these results shed light on guiding further exploration of cryptocurrency price forecasting in terms of the state-of-art neural networks.
A distributed ledger system is known as a block chain.Due to the frequent price changes of cryptocurrencies, they are by nature erratic.Hence, a platform was taken into consideration to monitor the growth of cryptocurrencies.The programme will keep track on bitcoin activity and provide information on value changes.In addition to leveraging an API to get the bitcoin data, we used several of the most well-known programming languages, like Python, to build the platform.The platform we created offers us data on the performance of crypto-currencies and has an intuitive user interface.The price changes from the previous day and the prior week are part of the daily updates to the cryptocurrency data that we get.A component of this is the value of cryptocurrencies.Making it simple for consumers to obtain cryptocurrency information was the main driver behind the development of this platform.Users may move between all of the pages with ease and no hassle thanks to the user interface (UI) we created.
Yakın geçmişte hayatımıza giren ve kısa zamanda finansal piyasalarda kendisine yer bulan kripto paralar, hem bir değişim aracı hem de bir yatırım aracı olarak kullanılmaktadır. Kripto paraların merkezi bir otoritenin kontrolünde olmaması bu araçların fiyatlarında dalgalanmaları beraberinde getirmiştir. Bu nedenle, akıllı bir tahmin modelinin geliştirilmesi, yatırım yapılacak finansal varlıkların seçimi ve yatırım kararlarının hayata geçirilmesi açısından oldukça önemlidir. Derin öğrenme ve yapay zeka, yatırım yapılacak olan kripto para birimi ve diğer yatırım araçlarının seçiminde kullanılmaktadır. Tekrarlayan Sinir Ağı (RNN), Uzun-Kısa Süreli Bellek (LSTM) ve Geçitli Yinelenen Birim (GRU) modeli gibi derin öğrenme modellerinin, kripto para birimi fiyat tahmininde geleneksel zaman serisi modellerinden daha iyi performans gösterdiği araştırmacılar tarafından kanıtlanmıştır. Bundan dolayı bu çalışmada, özel bir RNN yöntemi olan LSTM ve GRU’dan yararlanılarak, günümüzde piyasa değeri ve işlem hacmi en yüksek olan kripto paralardan Bitcoin, Ethereum ve Ripple’ın 30 günlük fiyat tahmininde bulunulmuştur. Araştırmanın sonucunda her iki modelde de en iyi tahmin sonucunu Bitcoin vermiştir. İkinci en iyi tahmin sonucu Ripple, sonrasında ise Ethereum için bulunmuştur. Kullanılan yöntemler karşılaştırıldığında ise MAPE performans ölçütüne göre en iyi tahmin sonucuna Bitcoin ve Ripple için GRU, Ethereum için ise LSTM modeli ile ulaşılmıştır.
Cryptocurrencies are popular today even though they do not have a physical form with their high profit rates and increasing usage day by day. However, the volatility of cryptocurrencies is higher than physical currencies. These volatilities change with the effect of social media rather than changes in exchange rates of physical currencies. For this reason, in this study, using Twitter data, one of the most widely used social media tools, real-time analysis on the values of four cryptocurrencies with the highest market value and the change in the estimated success compared to classical approaches were examined. The basic steps of this study: Obtaining Twitter data and financial data, performing sentiment analysis using Twitter data, making predictions on MM-LSTM architecture. The approach is aimed to be a predictive method open to online learning. Various filter steps were applied to remove the effect of bot users on Twitter that could prevent the prediction performance on the created data set, and the effect of the method on accuracy rate was tried to be reduced by eliminating the activity of bot accounts.
Researchers put efforts into explanations of the momentum phenomenon and improvements of the momentum strategy since the emergence of momentum in 1993. Interested in anomalies appearing as exhibited in traditional asset markets, adequate studies are launched on the nascent phenomenon emergers in the last decade, the cryptocurrency market. Recent studies have shown that there is hardly any cross-sectional momentum in the cryptocurrency market. To explore the momentum anomaly additionally in the cryptocurrency market, this paper implemented a time-series momentum on cross-sectional winners for improvement. Previous studies have introduced detecting the turning point between long-term slow time-series factor and short-term fast time-series factor contributes to predicting the trend well. Furthermore, a threshold decided by a certain machine learning model suggests better performance. In this paper. A multilayer perceptron (MLP) is utilized to learn the weights of time-series factors. The combination of cross-sectional momentum and time-series momentum shows advantages and the MLP learned weighted strategy is preferable.