What question did this study set out to answer?

To compare the performance of diverse machine learning and deep learning architectures in predicting price trends across twelve different cryptocurrencies.

January 1, 2024Facta universitatis - series Electronics and EnergeticsOpen Access

A comprehensive comparative study of machine learning models for predicting cryptocurrency

View Full Paper

Ask AI

Bookmark

Authors

YÜYüksel Akay ÜnvanAnkara University CECansu ErgençAnkara University

Discussion

Member takes

Implication

Comparative study reveals divergent predictive performance across multiple cryptocurrencies, highlighting that tree-based and deep neural networks offer asset-specific advantages.

Key Points

To compare the performance of diverse machine learning and deep learning architectures in predicting price trends across twelve different cryptocurrencies.
Benchmarked 11 predictive models: Decision Trees, Random Forests, KNN, GBM, LightGBM, XGBoost, CatBoost, ANN, CNN, RNN, and LSTM.
Analyzed price data across 12 cryptocurrencies, including Bitcoin, Ethereum, BinanceCoin, Solana, Cardano, Dogecoin, Tron, Aave, Cosmos, Tether, USDCoin, and XRP.
Evaluated model predictive accuracy using Mean Squared Error (MSE), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE).
No single machine learning model uniformly outperformed all alternatives across every cryptocurrency evaluated.
Tree-based ensemble models, specifically XGBoost and Random Forests, demonstrated consistent robustness and strong predictive stability across different assets.
Deep learning architectures, including CNNs, RNNs, and LSTMs, achieved the highest relative accuracy for specific individual cryptocurrencies.

Cite This Study

Ünvan et al. (2024) studied this question.

synapsesocial.com/papers/6a0fd757fa36b6e053fcf5d0 https://doi.org/10.2298/fuee2401211u

View Full Paper

Ask AI

Bookmark

Implication

Comparative study reveals divergent predictive performance across multiple cryptocurrencies, highlighting that tree-based and deep neural networks offer asset-specific advantages.

A comprehensive comparative study of machine learning models for predicting cryptocurrency

Discussion

Member takes

Implication

Key Points

Cite This Study

A comprehensive comparative study of machine learning models for predicting cryptocurrency

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider