LSTM Model Predicts USD/BDT Exchange Rates
A study (published Mar 30, 2026) uses 2018–2023 Yahoo Finance data to forecast USD/BDT exchange rates with an LSTM achieving 99.449% accuracy and RMSE 0.9858. A Gradient Boosting Classifier produced a 40.82% profitable trade rate but resulted in a net loss of $20,653.25 across 49 trades in backtesting. The authors recommend adding sentiment and real-time indicators to improve trading performance.
Key Points
- 1Demonstrates LSTM forecasting USD/BDT with 99.449% accuracy and RMSE 0.9858 on 2018–2023 data.
- 2Highlights GBC directional backtest: 40.82% profitable trades but net loss $20,653.25 over 49 trades.
- 3Signals deep learning improves point forecasts but requires better trading rules and risk management integration.
Scoring Rationale
Published today, this preprint shows strong LSTM point-forecast metrics for USD/BDT but poor trading outcomes. Scored for relevance and usable results, reduced for single-currency scope, single-source preprint status, and limited backtest success.
Sources
Public references used for this report.
Practice with real Ad Tech data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Ad Tech problems
