What question did this study set out to answer?

The aim is to enhance multivariate time series forecasting by addressing limitations in existing decomposition methods.

May 11, 2026Scientific ReportsOpen Access

CTMNet: causal trend evolution and adaptive modulation for time series forecasting

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates improved prediction accuracy in time series forecasting, suggesting advancements in modeling techniques.

The aim is to enhance multivariate time series forecasting by addressing limitations in existing decomposition methods.
Proposed the Causal Trend Evolution and Adaptive Modulation Network (CTMNet) for trend-residual interaction.
Introduced the Causal Trend Encoder (CTE) using causal convolution for unidirectional trend evolution.
Developed Adaptive Trend Modulation Interaction (ATMI) to dynamically adjust residual features.
CTMNet demonstrated leading performance on 10 benchmark datasets for long- and short-term forecasting.
Achieved significantly higher prediction accuracy compared to 7 state-of-the-art models.
Maintained physical consistency in trend modeling while enhancing dynamic capabilities.

Wang et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

May 11, 2026Scientific ReportsOpen Access

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates improved prediction accuracy in time series forecasting, suggesting advancements in modeling techniques.

The aim is to enhance multivariate time series forecasting by addressing limitations in existing decomposition methods.
Proposed the Causal Trend Evolution and Adaptive Modulation Network (CTMNet) for trend-residual interaction.
Introduced the Causal Trend Encoder (CTE) using causal convolution for unidirectional trend evolution.
Developed Adaptive Trend Modulation Interaction (ATMI) to dynamically adjust residual features.
CTMNet demonstrated leading performance on 10 benchmark datasets for long- and short-term forecasting.
Achieved significantly higher prediction accuracy compared to 7 state-of-the-art models.
Maintained physical consistency in trend modeling while enhancing dynamic capabilities.

Wang et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark