A Machine Learning Approach For Bitcoin Forecasting
Stefano Sossi-Rojas 1, Gissel Velarde 1, Damian Zieba 2 · 2025
Read A Machine Learning Approach For Bitcoin Forecasting by Stefano Sossi-Rojas 1, Gissel Velarde 1, Damian Zieba 2 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-18206.txt.
Introduction
Stefano Sossi-Rojas, Gissel Velarde, and Damian Zieba test which data streams help machine learning predict Bitcoin's price direction, in a 2025 paper. Earlier work hinted that the closing price alone cannot carry a forecast. The team assembled a wider set of time series and searched for the subset that sharpens directional calls.
The daily open, high, and low prices did the work, with the low price contributing most. Non-price features tied to the Bitcoin network added almost nothing. The best setup paired a gated recurrent unit network, a sequence-learning model, with a plain baseline forecast in an ensemble. Its directional accuracy matched the best published results. The authors frame the lesson as focus: a few price streams, well combined, beat a flood of extras.
Catalog record
| CREATOR | Stefano Sossi-Rojas 1, Gissel Velarde 1, Damian Zieba 2 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-18206
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A Machine Learning Approach For Bitcoin Forecasting Stefano Sossi-Rojas 1, Gissel Velarde 1, Damian Zieba 2 Abstract Bitcoin is one of the cryptocurrencies that is gaining more popularity in recent years. Previous studies have shown that closing price alone is not enough to forecast stock market series. We introduce a new set of time series and demonstrate that a subset is necessary to improve directional accuracy based on a machine learning ensemble. In our experiments, we study which time series and machine learning algorithms deliver the best results. We found that the most relevant time series that contribute to improving directional accuracy are Open, High and Low, with the largest contribution of Low in combination with an ensemble of Gated Recurrent Unit network and a baseline forecast. The relevance of other Bitcoin-related features that are not price-related is negligible. The proposed method delivers similar performance to the state-of-the-art when observing directional accuracy. 1 Introduction Bitcoin price forecasting has been the aim of many studies in the literature over the years. Nevertheless, unexpected price movements, smaller and larger bubbles, and different short- and long-term trends keep this task an ongoing topic for research. In one of the recent studies exploring this area of research [ 4 ] , authors verify the performance of different machine-learning algorithms and mention the current state of the knowledge on Bitcoin forecasting. One aspect is that referring to [ 3 ] the price of Bitcoin is mainly driven by spot market rather than futures market. The other aspect is the division of approaches to forecasting into “Blockchain approach” and “Financial market’s approach”. The former is based on technical variables such as hash rates or mining difficulty, while the latter uses standard econometric variables such as stocks, bonds, or gold price. What can be added to that is that the other approach is to take variables related to investor sentiment such as Google trends data [ 8 ] , uncertainty indices (VIX, UCRY, see [ 10 ] ), or the Fear and Greed index which has not been yet much explored in the literature. In our study, we use the mix of features from those three approaches as well as the Open, High, Low and Close prices to predict the Bitcoin price using a novel approach. Based on that, we are able to verify which variables contribute to the forecast accuracy the most. In fact, the features with the largest contribution are only the price-related ones. The other variables, whether coming from “Blockchain approach”, “Financial markets approach” or “Sentiment approach” have a negligible impact in improving Bitcoin’s price performance. We use a method that explores features and machine learning algorithms for Bitcoin’s closing price prediction. On recent forecasting competitions, it was observed that machine learning models and hybrid approaches demonstrated superiority over alternative methods [ 11 ] . Therefore, in this study we evaluated the following machine learning algorithms: Long Short Term Memory (LSTM), Bidirectional Long Term Memory (BiLSTM), Gated Recurrent Unit (GRU), Bidirectional Gated Recurrent Unit (BiGRU), and Light Gradient Boosting Machine (LightGBM). In addition we use ensembling. The results were evaluated observing the predicted and actual Bitcoin closing price measuring Root Mean Squared Error (RMSE), Mean Squared Error (MSE), Mean Absolute Error (MAE) and Directional Accuracy (DA). Previous studies show that stock market price is not enough for prediction when training deep learning models [ 19 ] , [ 18 ] . In this work, we demonstrate that indeed Bitcoin closing price is not sufficient for forecasting and additional features are necessary when using machine learning algorithms. Evaluation on return, shows that the method developed in this work presents one of the highest scores with a directional accuracy score equal to 0.7645 exceeding a baseline by 58.24 percent. We compare our work to related studies for Bitcoin prediction. In [ 1 ] , LSTM was used in combination to the Empirical Wavelet Transform (EWT) decomposition technique. The authors used the Intrinsic Mode Function (IMF) to optimize and estimate outputs with Cuckoo Search (CS) [ 1 ] . In [ 5 ] , Linear Regression (LR) techniques and particle swarm optimization were used to train and forecast data from beginning of 2012 to the end of March 2020. The Best setup for the model was obtained with 42 days plus 1 standard deviation [ 5 ] . In [ 21 ] , Autoregressive Integrated Moving Average (ARIMA) was used for data from 1 May 2013 to 7 June 2019. This model works best for short term predictions and can be used to predict Bitcoin for one to seven days ahead [ 21 ] . Finally, in [ 9 ] , a BiLSTM with Low-Middel-High features (LMH-BiLSTM) was tested with two primary steps: data decomposition and bidirectional deep learning. Results demonstrate that the proposed model outperforms other benchmark models and achieved high investment returns in the buy-and-hold strategy in a trading simulation [ 9 ] . In this work, we tested the previously mentioned machine learning algorithms one by one, and also in ensemble. We feed the algorithms with a new set of 13 time series, see section 2.1 . In addition, we included 11 signals that come from Variational Mode Decomposition (MMD) as proposed in [ 9 ] . However, in our experiments, data decomposition did not provide significant improvements. Next, we explain our proposed method. 2 Method Figure 1: Visual summary of the method. An overview of the method is presented in Fig. 1 . The input data is prepared in three steps. First, it is normalized between 0 and 1. Then, the train and test set partitions are created, where the train set is used for training several machine learning algorithms. A next phase considers algorithms’ training. The details on hyper-parameter selection are described in section 2.3 . Next, the evaluation phase is performed as rolling forecast for 1-step ahead prediction over the test set. 2.1 Data Collection We collected daily Bitcoin closing price from 7 October 2013 to 6 November 2022. We used a public API from the Kraken page [ 7 ] , the data collected are the values of: Close, Open, High, Low, Volume and Date. With the Nasdaq-Data-Link library for Python [ 12 ] the values were obtained: Transaction fee, Estimated Bitcoin USD Transaction Volume, Bitcoin USD Exchange Trade Volume, Bitcoin Hash Rate. Bitcoin Google trends were obtained with the pytrends library [ 14 ] . The Gold to USD Exchange Rate was obtained from the Investing.com page. The Fear and Greed Index was obtained from the Kaggle page [ 2 ] . The moving average of the Closing value was added, taking the last 30 days. Table 1 presents all features used in this study. A similar set of features has been used in [ 9 ] : Bitcoin Price, Bitcoin Transaction fees as Bitcoin miner’s revenue divided by transactions, USD trade volume from the top Bitcoin exchanges, Bitcoin transaction volume, USD exchange or trade volume from the top Bitcoin exchanges, Gold exchange rate to US dollar, Hash rate, and Google Trends of Bitcoin. Table 1: Series, Short Name, Description, and Count. Originally some signals had more samples than others, therefore, for those dates where no information was recorded