Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
DBLoss: Decomposition-based Loss Function for Time Series Forecasting
Authors: Xiangfei Qiu, Xingjian Wu, Hanyin Cheng, Xvyuan Liu, Chenjuan Guo, Jilin Hu, Bin Yang
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments demonstrate that DBLoss significantly improves the performance of state-of-the-art models across diverse real-world datasets and provides a new perspective on the design of time series loss functions. |
| Researcher Affiliation | Academia | Xiangfei Qiu1, Xingjian Wu1, Hanyin Cheng1, Xvyuan Liu1 Chenjuan Guo1, Jilin Hu1,2 , Bin Yang1 1East China Normal University, 2KLATASDS-MOE EMAIL, EMAIL |
| Pseudocode | Yes | Algorithm 1 Calculation Process of EMA Decomposition Module |
| Open Source Code | Yes | Resources: https://github.com/decisionintelligence/DBLoss. |
| Open Datasets | Yes | Datasets To conduct comprehensive and fair comparisons for different models, we conduct experiments on eight well-known forecasting benchmarks as the target datasets, including ETT (ETTh1, ETTh2, ETTm1, ETTm2), Solar, Weather, Electricity, and Traffic. For more details on the benchmark datasets, please refer to Table 5 in Appendix A. |
| Dataset Splits | Yes | For more details on the benchmark datasets, please refer to Table 5 in Appendix A. Table 5: Statistics of datasets. Dataset Domain Frequency Lengths Dim Split Description ... ETTh1 Electricity 1 hour 14,400 7 6:2:2 Power transformer 1, comprising seven indicators such as oil temperature and useful load |
| Hardware Specification | Yes | All experiments of DBLoss are conducted using Py Torch in Python 3.8 and executed on an NVIDIA Tesla-A800 GPU. |
| Software Dependencies | Yes | All experiments of DBLoss are conducted using Py Torch in Python 3.8 and executed on an NVIDIA Tesla-A800 GPU. |
| Experiment Setup | Yes | We consider four forecasting horizon F: {96, 192, 336, 720} for all datasets. The training process is guided by the MSE loss function and employs the ADAM optimizer. The initial batch size is set to 64, with the flexibility to halve it (down to a minimum of 8) in case of an Out-Of-Memory (OOM) issue. |