Harnessing Loss Decomposition for Long-Horizon Wave Predictions via Deep Neural Networks

TL;DR

This paper introduces a loss decomposition method that separates phase and amplitude errors, significantly enhancing the long-term accuracy of neural network predictions in wave propagation tasks by reducing error accumulation.

Contribution

The paper presents a novel loss decomposition strategy that explicitly separates phase and amplitude errors to improve long-horizon wave predictions with neural networks.

Findings

Enhanced long-term wave prediction accuracy.

Reduced error accumulation over extended forecasts.

Improved stability in neural network-based wave modeling.

Abstract

Accurate prediction over long time horizons is crucial for modeling complex physical processes such as wave propagation. Although deep neural networks show promise for real-time forecasting, they often struggle with accumulating phase and amplitude errors as predictions extend over a long period. To address this issue, we propose a novel loss decomposition strategy that breaks down the loss into separate phase and amplitude components. This technique improves the long-term prediction accuracy of neural networks in wave propagation tasks by explicitly accounting for numerical errors, improving stability, and reducing error accumulation over extended forecasts.