[PDF] Forex exchange rate forecasting using deep recurrent neural networks | Semantic Scholar (2024)

The TLS-LSTM algorithm is presented to improve the accuracy of Forex trend prediction of Australian Dollar and United States Dollar (AUD/USD) and outperforms SL-LSTM, MLP, and CEEMDAN-IFALSTM.

This paper compares forecasting models to evaluate their accuracy in the short term using data on the EUR/USD exchange rate and finds that the window of one month with 22 observations better matched the validation dataset in theshort term compared to the other windows.

This paper proposes an early warning system for currency crises using sophisticated recurrent neural networks, such as long short‐term memory (LSTM) and gated recurrent unit (GRU), which outperformed the authors' benchmark models.

This research used research efforts on LM, MLP, and Recurrent Neural Networks (RNNs) for predicting EUR/USD FOREX rates and MSE was used for evaluating the result.

A two-stage predictive model that combines regression and classification tasks, using the predicted closing price to determine entry and exit points is developed, demonstrating effectiveness and reliability of the AMBO-DSN approach in forecasting trends for USD/EUR, AUD/JPY, and CHF/INR currency pairs.

This paper uses technical indicators as features of historical data, fundamental analysis to collect detailed and comprehensive external information, and Ensemble Feature Grader (EFG) to filter out the noise in the features.

A fundamental-based model is developed using machine learning to predict the exchange rate and interpretability methods to accurately analyze the relationships among macroeconomic variables and an ablation study is implemented based on the output of the interpretations to improve the predictive accuracy of the models.

The authors develop a model for a near-optimal state trying to avoid uncertainty and insufficient accuracy of the VIX and make a contribution to the theory of socially responsible portfolio management.

...

...

The application of DNNs to predicting financial market movement directions is described and their application to back testing a simple trading strategy over 43 different Commodity and FX future mid-prices at 5-minute intervals is demonstrated.

The use of recurrent neural networks in order to forecast foreign exchange rates is proposed and the methods described here which have obtained promising results in real time trading are applicable to other markets.

A Long Short-Term Memory neural network is applied to model S&P 500 volatility, incorporating Google domestic trends as indicators of the public mood and macroeconomic factors and shows strong promise for better predicting stock behavior via deep learning and neural network models.

The results show that PSC is a sensible criterion for selecting networks and for certain exchange rate series, some selected network models have significant market timing ability and/or significantly lower out-of-sample prediction error relative to the random walk model.

It is shown that the symbolic representation aids the extraction of symbolic knowledge from the trained recurrent neural networks in the form of deterministic finite state automata which explain the operation of the system and are often relatively simple.

The results for forecast evaluation for all the models for each of the series based on summary measures of forecast evaluations show that RNN3 model appears to offer the most accurate predictions of BP and RNN1 for JP exchange rates, however, none of the RNN models appear to be statistically superior to the benchmark for predicting CD exchange rates.

...

...