EUVIMEDEUVIMED entdecken
Crossref

Hybrid Transformer–LSTM for Accurate AAVE Price Forecasting Using Technical Indicators in Decentralized Finance Ecosystems through Deep Learning Models

Guevara Ananta Toer

Journal of Current Research in Blockchain · 02.11.2026 · 3 (4) · 298-314

DOI: 10.47738/jcrb.v3i4.86 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Current Research in Blockchain, online erschienen am 02.11.2026. Verfasst von Guevara Ananta Toer.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The rapid growth of decentralized finance (DeFi) has increased the importance of accurate cryptocurrency price forecasting to support intelligent financial decision making in emerging metaverse ecosystems. However, the high volatility and nonlinear characteristics of cryptocurrency markets make reliable prediction a challenging task. This study proposes a Hybrid Transformer LSTM model for forecasting AAVE cryptocurrency prices using historical price data and technical indicators. Daily OHLC data collected from November 2020 to March 2026 were employed to compute multiple technical indicators, including Simple Moving Average, Exponential Moving Average, Relative Strength Index, Moving Average Convergence Divergence, Average True Range, Rate of Change, Momentum, Bollinger Bands, Williams %R, Commodity Channel Index, and Stochastic Oscillator. The proposed model was compared with four benchmark approaches, namely the Naïve model, LSTM, GRU, and Transformer, using Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), coefficient of determination (R²), and Directional Accuracy (DA). Experimental results show that all deep learning models successfully captured the general trend of AAVE price movements and achieved stable convergence during training. The Naïve model obtained the lowest prediction errors and the highest R², while the GRU achieved the highest Directional Accuracy. Although the proposed Hybrid Transformer LSTM did not achieve the best numerical performance, it produced stable prediction trajectories and a balanced residual distribution, demonstrating consistent forecasting behavior under different market conditions. These findings indicate that hybrid attention and recurrent learning architectures provide a robust framework for cryptocurrency forecasting, while also highlighting that increased model complexity does not necessarily lead to superior predictive accuracy when only technical indicators are used. The study contributes to intelligent financial analytics for decentralized finance and metaverse applications by providing a comprehensive comparison of contemporary deep learning architectures for cryptocurrency price prediction.

Zitieren

APA

Toer, G. A. (2026). Hybrid Transformer–LSTM for Accurate AAVE Price Forecasting Using Technical Indicators in Decentralized Finance Ecosystems through Deep Learning Models. Journal of Current Research in Blockchain, 3(4), 298-314. https://doi.org/10.47738/jcrb.v3i4.86

Vancouver

Toer GA. Hybrid Transformer–LSTM for Accurate AAVE Price Forecasting Using Technical Indicators in Decentralized Finance Ecosystems through Deep Learning Models. Journal of Current Research in Blockchain. 2026;3(4):298-314. doi:10.47738/jcrb.v3i4.86

BibTeX
@article{Toer2026,
  author = {Toer, Guevara Ananta},
  title = {Hybrid Transformer–LSTM for Accurate AAVE Price Forecasting Using Technical Indicators in Decentralized Finance Ecosystems through Deep Learning Models},
  journal = {Journal of Current Research in Blockchain},
  year = {2026},
  volume = {3},
  number = {4},
  pages = {298--314},
  doi = {10.47738/jcrb.v3i4.86}
}
Kostenlos starten →