Adaptive Temporal Fusion Transformers for Cryptocurrency Price Prediction

ArXiv ID: 2509.10542 “View on arXiv”

Authors: Arash Peik, Mohammad Ali Zare Chahooki, Amin Milani Fard, Mehdi Agha Sarram

Abstract

Precise short-term price prediction in the highly volatile cryptocurrency market is critical for informed trading strategies. Although Temporal Fusion Transformers (TFTs) have shown potential, their direct use often struggles in the face of the market’s non-stationary nature and extreme volatility. This paper introduces an adaptive TFT modeling approach leveraging dynamic subseries lengths and pattern-based categorization to enhance short-term forecasting. We propose a novel segmentation method where subseries end at relative maxima, identified when the price increase from the preceding minimum surpasses a threshold, thus capturing significant upward movements, which act as key markers for the end of a growth phase, while potentially filtering the noise. Crucially, the fixed-length pattern ending each subseries determines the category assigned to the subsequent variable-length subseries, grouping typical market responses that follow similar preceding conditions. A distinct TFT model trained for each category is specialized in predicting the evolution of these subsequent subseries based on their initial steps after the preceding peak. Experimental results on ETH-USDT 10-minute data over a two-month test period demonstrate that our adaptive approach significantly outperforms baseline fixed-length TFT and LSTM models in prediction accuracy and simulated trading profitability. Our combination of adaptive segmentation and pattern-conditioned forecasting enables more robust and responsive cryptocurrency price prediction.

Keywords: Temporal Fusion Transformers (TFT), Cryptocurrency Forecasting, Pattern-Based Categorization, Adaptive Segmentation, Non-Stationary Time Series, Cryptocurrencies

Complexity vs Empirical Score

• Math Complexity: 7.0/10
• Empirical Rigor: 5.5/10
• Quadrant: Holy Grail
• Why: The paper employs advanced deep learning architectures (Temporal Fusion Transformers) with specialized algorithms (adaptive segmentation and pattern-conditioned categorization), indicating significant mathematical complexity. It also provides empirical validation on real cryptocurrency data with trading simulations, demonstrating solid implementation and data-heavy experimentation.

  flowchart TD
    A["Research Goal: Enhance Short-Term Cryptocurrency Price Prediction in Volatile Markets"] --> B["Data Input: ETH-USDT 10-Minute Price Data"]
    B --> C["Adaptive Segmentation Strategy<br>Split data at relative maxima<br>Capture key growth phases"]
    C --> D["Pattern-Based Categorization<br>Group subseries by preceding fixed-length patterns"]
    D --> E["Category-Specific TFT Modeling<br>Train distinct TFT models per category"]
    E --> F["Computational Outcome: Multi-Model Forecasting System"]
    F --> G["Key Findings: Superior Accuracy & Trading Profitability<br>Outperforms Fixed-Length TFT & LSTM Baselines"]