Comparative study of Bitcoin price prediction using WaveNets, Recurrent Neural Networks and other Machine Learning Methods

Felizardo, Leonardo Kanashiro; Oliveira, R. A.; Del-Moral-Hernandez, Emílio; Cozman, Fábio Gagliardi

doi:10.1109/besc48373.2019.8963009

Cited by 17 publications

(8 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Different methods were applied in the construction of the Bitcoin price prediction model to build a reliable model, which is contrasted with various methodologies used in previous works to check with which technique a high predictive capacity is achieved; specifically, the methods of deep recurrent neural networks, deep neural decision trees, and deep support vector machines, were used. Furthermore, this work attempts to obtain high accuracy, but it is also robust and stable in the future horizon to predict new observations, something that has not yet been reported by previous works [7][8][9][10][11][12][13][14][15], but which some authors demand for the development of these models and their real contribution [9,12].…”

Section: Introductionmentioning

confidence: 94%

“…Linardatos and Kotsiantis [12] had the same results, after using eXtreme Gradient Boosting (XGBoost) and LSTM; they concluded that this last technique yielded a lower RMSE of 0.999. Despite the superiority of computational techniques, Felizardo and colleagues [13] showed that ARIMA had a lower error rate than methods, such as random forest (RF), support vector machine (SVM), LSTM, and WaveNets, to predict the future price of Bitcoin. Finally, other works showed new deep learning methods, such as Dutta, Kumar, and Basu [14], who applied both LSTM and the gated recurring unit (GRU) model; the latter showed the best error result, with an RMSE of 0.019.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep Learning Methods for Modeling Bitcoin Price

et al. 2020

View full text Add to dashboard Cite

A precise prediction of Bitcoin price is an important aspect of digital financial markets because it improves the valuation of an asset belonging to a decentralized control market. Numerous studies have studied the accuracy of models from a set of factors. Hence, previous literature shows how models for the prediction of Bitcoin suffer from poor performance capacity and, therefore, more progress is needed on predictive models, and they do not select the most significant variables. This paper presents a comparison of deep learning methodologies for forecasting Bitcoin price and, therefore, a new prediction model with the ability to estimate accurately. A sample of 29 initial factors was used, which has made possible the application of explanatory factors of different aspects related to the formation of the price of Bitcoin. To the sample under study, different methods have been applied to achieve a robust model, namely, deep recurrent convolutional neural networks, which have shown the importance of transaction costs and difficulty in Bitcoin price, among others. Our results have a great potential impact on the adequacy of asset pricing against the uncertainties derived from digital currencies, providing tools that help to achieve stability in cryptocurrency markets. Our models offer high and stable success results for a future prediction horizon, something useful for asset valuation of cryptocurrencies like Bitcoin.

show abstract

Section: Introductionmentioning

confidence: 94%

Section: Introductionmentioning

confidence: 99%

Deep Learning Methods for Modeling Bitcoin Price

et al. 2020

View full text Add to dashboard Cite

show abstract

“…In addition, RF can work faster and better on large training samples, with the training time significantly faster than the SVM. However, some studies have shown that RF performed worse than the regression models in modeling linear data [55,113]. In short, it is suitable for nonlinear, stationary, and highly volatile data.…”

Section: Support Vector Machine (Svm)mentioning

confidence: 99%

Exploring Time-Series Forecasting Models for Dynamic Pricing in Digital Signage Advertising

et al. 2021

View full text Add to dashboard Cite

Audience attention is vital in Digital Signage Advertising (DSA), as it has a significant impact on the pricing decision to advertise on those media. Various environmental factors affect the audience attention level toward advertising signage. Fixed-price strategies, which have been applied in DSA for pricing decisions, are generally inefficient at maximizing the potential profit of the service provider, as the environmental factors that could affect the audience attention are changing fast and are generally not considered in the current pricing solutions in a timely manner. Therefore, the time-series forecasting method is a suitable pricing solution for DSA, as it improves the pricing decision by modeling the changes in the environmental factors and audience attention level toward signage for optimal pricing. However, it is difficult to determine an optimal price forecasting model for DSA with the increasing number of available time-series forecasting models in recent years. Based on the 84 research articles reviewed, the data characteristics analysis in terms of linearity, stationarity, volatility, and dataset size is helpful in determining the optimal model for time-series price forecasting. This paper has reviewed the widely used time-series forecasting models and identified the related data characteristics of each model. A framework is proposed to demonstrate the model selection process for dynamic pricing in DSA based on its data characteristics analysis, paving the way for future research of pricing solutions for DSA.

show abstract

“…The price time series data is processed into tensors and fed into a LSTM model, which achieved a better F1-score for this binary classification task, compared to a baseline gradient boosting model. LSTMs are a popular architecture for time-series prediction, and other researchers have also used them to predict cryptocurrency prices using technical indicators (Aditya Pai, Devareddy, Hegde and Ramya, 2022), (Wu, Lu, Ma and Lu, 2018) (Shin, Mohaisen and Kim, 2021), (Felizardo, Oliveira, Del-Moral-Hernandez and Cozman, 2019). Using a different neural network architecture, Alonso-Monsalve, Suárez-Cetrulo, Cervantes and Quintana (2020) adopted Convolutional Neutral Networks (CNNs) and compared their prediction results to a simple Multilayer Perceptron (MLP), as well as an LSTM model.…”

Section: Btc Price Prediction With Technical Indicatorsmentioning

confidence: 99%

“…WaveNet is a network based upon the concept of Temporal CNNs with residual connections. Felizardo et al (2019) did a comparison study between WaveNet, ARIMA, Random Forest, SVM, and LSTM to predict the future price of Bitcoin using OHLCV data. They found that the ARIMA model and the SVM equally outperformed the other models including WaveNet.…”

Section: Btc Price Prediction With Technical Indicatorsmentioning

confidence: 99%

A multimodal model with Twitter FinBERT embeddings for extreme price movement prediction of Bitcoin

Zou¹,

Herremans²

2022

Preprint

View full text Add to dashboard Cite

Proposes a multimodal model for Bitcoin extreme price movement prediction using Twitter contents, not just sentiment.• Ablation study of the impact of different modalities on extreme price prediction accuracy.• New publicly available dataset of 9,435,437 tweets over a period of years related to Bitcoin.• Describes a profitable strategy with reduced risk exposure for Bitcoin trading using the proposed multimodal model.• Demonstrates the influence of setting predictive thresholds on risk reduction of a trading strategy.

show abstract

Comparative study of Bitcoin price prediction using WaveNets, Recurrent Neural Networks and other Machine Learning Methods

Cited by 17 publications

References 13 publications

Deep Learning Methods for Modeling Bitcoin Price

Deep Learning Methods for Modeling Bitcoin Price

Exploring Time-Series Forecasting Models for Dynamic Pricing in Digital Signage Advertising

A multimodal model with Twitter FinBERT embeddings for extreme price movement prediction of Bitcoin

Contact Info

Product

Resources

About