Paper: arXiv 2411.05829
Authors: Shamima Nasrin Tumpa, Kehelwala Dewage Gayan Maduranga
Abstract
This study explores the use of Recurrent Neural Networks (RNN) for real-time cryptocurrency price prediction and optimized trading strategies. Given the high volatility of the cryptocurrency market, traditional forecasting models often fall short. By leveraging RNNs’ capability to capture long-term patterns in time-series data, this research aims to improve accuracy in price prediction and develop effective trading strategies. The project follows a structured approach involving data collection, preprocessing, and model refinement, followed by rigorous backtesting for profitability and risk assessment. This work contributes to both the academic and practical fields by providing a robust predictive model and optimized trading strategies that address the challenges of cryptocurrency trading.
Complexity vs Empirical Score
- Math Complexity: 6.5/10
- Empirical Rigor: 7.0/10
- Quadrant: Holy Grail — high math complexity, high empirical rigor
Why this score: The paper employs advanced neural network architectures (LSTM, GRU, Bi-LSTM) and rigorous time-series forecasting methodology with standard metrics, and includes a detailed backtesting framework on real cryptocurrency data.
Research Flowchart
flowchart TD A["Research Goal: Improve crypto price prediction & trading strategy using RNN"] --> B["Data Collection<br>Historical Crypto Price Data"] B --> C["Preprocessing & Feature Engineering<br>Normalization, Time-series Structuring"] C --> D["Model Building: RNN Architecture<br>LSTM/GRU for long-term pattern capture"] D --> E["Training & Validation<br>Optimize parameters on historical data"] E --> F["Strategy Development<br>Generate trading signals from predictions"] F --> G["Backtesting & Evaluation<br>Simulate trading, assess profitability & risk"] G --> H["Key Outcomes<br>Robust predictive model & optimized trading strategy"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a robust predictive model and optimized trading strategies for cryptocurrency price prediction.
Source sentence
This work contributes to both the academic and practical fields by providing a robust predictive model and optimized trading strategies that address the challenges of cryptocurrency trading.
Assumes 1
- Missing price observations are imputed by carrying forward the last available observation. data
Source sentence
missing data points were replaced with the last available observation.
Data
- Bitcoin, Ethereum, Litecoin historical price data market-prices unknown
Source sentence
The process begins with gathering datasets for Bitcoin (BTC), Ethereum (ETH), and Litecoin (LTC)
Yahoo Finance
Costs
Not applicable
Finds 2
- Bi-LSTM achieved the lowest error for Bitcoin with MSE 0.0001237, MAE 0.007581, RMSE 0.011. performance
Source sentence
For BTC, the Bi-LSTM model achieved the best result, demonstrating its capability to capture complex patterns in the price data with an MSE of 0.0001237, MAE of 0.007581, and RMSE of 0.011.
- GRU outperformed other models for Ethereum and Litecoin, achieving the smallest errors (e.g., ETH MSE 0.0000905, LTC MSE 0.0000730). performance
Source sentence
the GRU model outperformed the other models for both ETH and LTC, achieving the smallest errors. For ETH, GRU recorded an MSE of 0.0000905, MAE of 0.006697, and RMSE of 0.010, while for LTC, it achieved an MSE of 0.0000730, MAE of 0.006096, and RMSE of 0.009.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 17 related papers). Coverage is partial, so this is neither replication nor refutation.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.