Aug 2026· Artificial Intelligence Review· 0 citations
TL;DR
This survey re-examines deep learning models for MTS forecasting through the requirements of efficiency and explainability, and identifies key open challenges including the absence of standardized explainability benchmarks for time series, the interpretability gap in state space models, and the need to advance from correlational to causal explanations.
Abstract
The rapid expansion of multivariate time series (MTS) data has made deep learning a central tool for forecasting across industrial and scientific domains. As these architectures move into settings such as clinical decision support, industrial monitoring, and financial risk management, researchers need to examine not only predictive accuracy but also computational cost and the kinds of explanations that a model can support. This survey re-examines deep learning models for MTS forecasting through the requirements of efficiency and explainability. Our contributions are threefold: (1) We synthesize the literature through a dual-axis scalability view over sequence length
L
and variate dimension
N
, together with a three-question explainability taxonomy covering temporal importance, variate importance, and pattern decomposition. (2) We review dominant architectural families, including Transformer variants, frequency-domain methods, linear and channel-independent backbones, state space models, and foundation-style forecasters, to describe how different designs occupy different efficiency–explainability profiles under specific data and deployment assumptions. (3) Building on this synthesis, we identify key open challenges—including the absence of standardized explainability benchmarks for time series, the interpretability gap in state space models, and the need to advance from correlational to causal explanations—and offer reporting considerations to guide more comparable future research.
Forecasting time series over long horizons is essential for proactive decision-making in many systems. Recent research has focused on transformer-based architectures, which capture long-range dependencies in sequential data. However, several studies show that simpler linear models can outperform transformers by avoiding overfitting during training. In this context, we present NeuroFlexMLP, a deep learning model for multivariate time series forecasting tasks. NeuroFlexMLP's key distinct feature is the adaptability to the diverse complexity of real-world time series, which is achieved, from the architecture standpoint, by adding non-linear residual blocks to a first linear block. This architectural design simplifies hyperparameter optimization, leading to accurate forecasts for various time series data types regardless of the lookback or prediction horizons, outperforming state-of-the-art (SOTA) models on challenging real-world datasets. Its Multi-Layer Perceptron (MLP) design ensures high computational efficiency, making it scalable for longer input sequences than transformer-based models. We validate NeuroFlexMLP for the LEO satellite beam hopping use case, where its lightweight design enables on-board deployment, and on state-of-the art AI datasets. Across all these benchmarks, NeuroFlexMLP achieves competitive accuracy over state-of-the-art models while providing an adaptive architecture that significantly reduces computational overhead. On the LEO beam hopping task, it achieves up to 35.9% MSE reduction over Informer, which translates into up to 28% lower provisioning cost under asymmetric cost models that penalize under-allocation more heavily than over-allocation.
P. F. Pérez, Claudio Fiandrino, Marco Fiore et al.· La Main· 0 citations
The literature is organized around five recurring difficulties: nonlinear and nonstationary behavior, contamination and structural breaks, uncertainty, long contexts and cross-variable dependence, and limited target-domain data.
Chu-Ting Wen· Applied and Computational En...· 0 citations
Multivariate Time Series Classification (MTSC) is a central task in modern data analytics, with growing impact across domains such as healthcare, finance, and industrial monitoring. As MTSC models are increasingly used in real-world decision-making, the need for explainability has become critical. Existing solutions either rely on feature-extraction frameworks that produce opaque descriptors or on explainable-by-design models tied to specific architectures. We introduce EFFECTS, a scalable framework for extracting interpretable temporal features that enable model-agnostic explanations of black-box classifiers and also support the construction of inherently interpretable models. EFFECTS automatically identifies meaningful time slices and characterizes how they distinguish between classes through intuitive transformations and aggregations. Extensive experiments, including a user study with both novice and experienced data scientists, show that EFFECTS produces clearer and more actionable explanations than architecture-specific methods, while maintaining competitive accuracy and strong runtime performance. The source code and data have been made available at https://github.com/analysis-bots/EFFECTS.
Ido Ikar, Amit Somech· Proceedings of the 32nd ACM...· 0 citations
This study compares ARIMA, LSTM, and temporal fusion transformer (TFT) models across three applications and shows that TFT consistently achieved superior forecasting performance and demonstrated greater robustness to increasing missingness, while k-NN generally provided the most effective imputation performance across datasets.
M. Hosseini, Mohamad Forouzanfar· Computer Science and Informa...· 0 citations
The findings demonstrate that rigorous leakage-free validation is essential for reliable forecasting research and that, for monthly Robusta coffee prices, increased model complexity does not necessarily yield superior predictive performance.
Dler H Kadir, D. Khalil, Azhin M. Khudhur· Forecasting· 0 citations
Accurate prediction of financial time series is still a difficult problem as financial markets display high volatility, non-linearity and stochasticity. Traditional forecasting methods necessitate extensive domain knowledge in designing technical indicators for subsequent analysis, often resulting in the loss of intricate time dependencies. The goal of the present study is to propose a framework allowing for learning representations automatically from raw financial data that are informative in downstream forecasting tasks. The proposed framework, contrasting predictive coding (CPC), is based on self-supervised representation learning. The learned embeddings are applied to Linear Regression, Random Forest and LSTM to predict the next-day log returns of three major foreign exchange currency pairs: EUR/USD, GBP/USD and USD/JPY. Evaluating the Performance of CPC-Generated Representations and Conventional Handcrafted Features on Forecasting Models trained on Historical Market Data. The LSTM with CPC context embeddings produces the best overall performance with a drop in mean squared error of 18%, directional prediction accuracy of roughly 59%, and better risk-adjusted trading performance with Sharpe ratios above 0.7. Additionally, the outcomes of transfer learning experiments reveal that a CPC encoder trained using one currency pair efficiently generalizes to other currency pairs. The results indicate that self-supervised representation learning can serve as an effective and scalable substitute for manual feature engineering in finance time-series forecasting.
Muskan Pawar· International Journal of Ada...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
