Beyond Holistic Models: Systematic Component-level Benchmarking of Deep Multivariate Time-Series Forecasting

• URL: http://arxiv.org/abs/2605.26562v1
• Date: Tue, 26 May 2026 05:18:38 GMT
• Title: Beyond Holistic Models: Systematic Component-level Benchmarking of Deep Multivariate Time-Series Forecasting
• Abstract summary: We propose TSCOMP, the first large-scale benchmark that systematically deconstructs deep forecasting methods into their core, fine-grained components.<n>We conduct multi-view analyses that reveal component effectiveness across different backbones, data characteristics, and their interactions.<n>Our experiments demonstrate that the corpus-driven approach consistently outperforms state-of-the-art methods.
• Score: 38.12202305030755
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: While previous research in multivariate time series forecasting has focused on developing complex holistic models, this work advocates for a shift toward a granular, component-level understanding of their impacts. We propose TSCOMP, the first large-scale benchmark that systematically deconstructs deep forecasting methods into their core, fine-grained components--spanning series preprocessing, encoding strategies, network architectures including specific and large time-series models, and optimization methods. Using constrained orthogonal experimental design and extensive evaluations, we conduct multi-view analyses that reveal component effectiveness across different backbones, data characteristics, and their interactions. Beyond providing insights, this benchmark establishes a fine-grained performance corpus comprising over 20,000 model-dataset evaluations, which supports the learning of automated component selection, enabling zero-shot model construction on new datasets. Our experiments demonstrate that the corpus-driven approach, despite its simplicity, consistently outperforms state-of-the-art methods, validating the soundness of our evaluation design and confirming that systematic component selection surpasses manually designed complex architectures. All code and the performance corpus are publicly available at https://github.com/SUFE-AILAB/TSCOMP.

Related papers

• Cognitive-YOLO: LLM-Driven Architecture Synthesis from First Principles of Data for Object Detection [3.5554162308775408]
  We propose Cognitive-YOLO, a novel framework for Large Language Models (LLMs)-driven architecture synthesis.<n>Our method consists of three stages: first, an analysis module extracts key meta-features from the target dataset.<n>Second, the LLM reasons upon these features, augmented with state-of-the-art components retrieved via Retrieval-Augmented Generation (RAG), to synthesize the architecture into a structured Neural Architecture Description Language (NADL)<n>Third, a compiler instantiates this description into a deployable model.
  arXiv  Detail & Related papers  (2025-12-13T10:52:54Z)
• ProtoTSNet: Interpretable Multivariate Time Series Classification With Prototypical Parts [6.99674326582747]
  ProtoTSNet is a novel approach to interpretable classification of time series data.<n>Central to our innovation is a modified convolutional encoder utilizing group convolutions, pre-trainable as part of an autoencoder.
  arXiv  Detail & Related papers  (2025-11-04T00:42:42Z)
• Segmentation over Complexity: Evaluating Ensemble and Hybrid Approaches for Anomaly Detection in Industrial Time Series [36.94429692322632]
  We evaluate the impact of change point-derived statistical features, clustering-based substructure representations, and hybrid learning strategies on detection performance.<n>The ensemble achieved an AUC-ROC of 0.976, F1-score of 0.41, and 100% early detection within the defined time window.
  arXiv  Detail & Related papers  (2025-10-30T05:39:44Z)
• TimeRecipe: A Time-Series Forecasting Recipe via Benchmarking Module Level Effectiveness [21.753566949357026]
  TimeRecipe is a framework that systematically evaluates time-series forecasting methods at the module level.<n>TimeRecipe conducts over 10,000 experiments to assess the effectiveness of individual components.<n>Our results reveal that exhaustive exploration of the design space can yield models that outperform existing state-of-the-art methods.
  arXiv  Detail & Related papers  (2025-06-06T19:11:48Z)
• ArtGS: Building Interactable Replicas of Complex Articulated Objects via Gaussian Splatting [66.29782808719301]
  Building articulated objects is a key challenge in computer vision.<n>Existing methods often fail to effectively integrate information across different object states.<n>We introduce ArtGS, a novel approach that leverages 3D Gaussians as a flexible and efficient representation.
  arXiv  Detail & Related papers  (2025-02-26T10:25:32Z)
• Embedded feature selection in LSTM networks with multi-objective evolutionary ensemble learning for time series forecasting [49.1574468325115]
  We present a novel feature selection method embedded in Long Short-Term Memory networks. Our approach optimize the weights and biases of the LSTM in a partitioned manner. Experimental evaluations on air quality time series data from Italy and southeast Spain demonstrate that our method substantially improves the ability generalization of conventional LSTMs.
  arXiv  Detail & Related papers  (2023-12-29T08:42:10Z)
• Revisiting the Evaluation of Image Synthesis with GANs [55.72247435112475]
  This study presents an empirical investigation into the evaluation of synthesis performance, with generative adversarial networks (GANs) as a representative of generative models. In particular, we make in-depth analyses of various factors, including how to represent a data point in the representation space, how to calculate a fair distance using selected samples, and how many instances to use from each set.
  arXiv  Detail & Related papers  (2023-04-04T17:54:32Z)
• Learning from Temporal Spatial Cubism for Cross-Dataset Skeleton-based Action Recognition [88.34182299496074]
  Action labels are only available on a source dataset, but unavailable on a target dataset in the training stage. We utilize a self-supervision scheme to reduce the domain shift between two skeleton-based action datasets. By segmenting and permuting temporal segments or human body parts, we design two self-supervised learning classification tasks.
  arXiv  Detail & Related papers  (2022-07-17T07:05:39Z)
• Few-Shot Named Entity Recognition: A Comprehensive Study [92.40991050806544]
  We investigate three schemes to improve the model generalization ability for few-shot settings. We perform empirical comparisons on 10 public NER datasets with various proportions of labeled data. We create new state-of-the-art results on both few-shot and training-free settings.
  arXiv  Detail & Related papers  (2020-12-29T23:43:16Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.