Overview

Time series forecasting studies how to predict future observations from past values, often across multiple horizons and with missing data, changing regimes, or external drivers such as weather and events. Methods range from classical autoregressive and exponential-smoothing models to recurrent networks, Transformers, state-space models, decomposition and frequency-based approaches, graph models, and pretrained foundation models; probabilistic methods also estimate uncertainty. Current work increasingly combines historical patterns with covariates, retrieval, domain knowledge, and online adaptation, while targeting applications from energy and weather to health, transport, and finance. A central tension is that larger or more complex models do not reliably beat simple baselines, and benchmark accuracy may not reflect robustness, forecast structure, uncertainty calibration, or downstream decision value.

Latest papers

947 papers84 papers in the last four weeks, down 49% on the four weeks before. 0.6% of all new papers.

Nov 19, 2026

The Window Lever: Longer Observation Windows Outweigh the Observed Effects of Data, Capacity, and Architecture in Customer Behavior Forecasting

In our experiments on long-horizon customer-behavior forecasting, a longer observation window outweighed the observed effects of more data, more capacity and newer architectures. The window is the weeks of history a model sees before the week it forecasts from. Going from an 8- to a 16-week window adds a raw +0.070 to 0.107 log-space R2R^2 in every family at 1M. Matched test cohorts, which score the same customer-weeks at every window, show this raw comparison understates the gain in every family they cover. The direction holds on five public datasets, four spanning over a year (ridge and our transformer on all five). On our internal data, longer windows admit up to 10% more and colder training customers, which matching does not fix. Our design does not separate window length from calendar coverage, because every test anchor (the week a forecast is made from) falls in one eight-week band. On training-period anchors, the weekly ridge probe's gain persists outside the holiday season, keeping about 72% of it. We therefore report a robust positive association with longer windows and do not identify the magnitude of a calendar-independent history effect. Of four internally registered predictions, the only fully blind one failed. An activity-mixture model predicted +0.014 from 24 to 32 weeks; the weekly-aggregate probe returned -0.0004. For the sequence families, the other three levers are about an order of magnitude smaller, bounded and unresolved. Their observed effects are at most 0.006 between the nominal 1M and 10M tiers and 0.0052 among five sequence families at 16 weeks with clocks as built, both single-seed on the full test split, and 0.006 for capacity on three-seed means. These are observed intervals, not equivalence bounds. Our claim is the order-of-magnitude contrast between the window and those bounds. A team building on this class of data should start by lengthening the window.
Nov 19, 2026

Interpretable Prototype-Conditioned Adaptation for Weather Forecasting

Data-driven weather forecasting requires distinguishing local states and their temporal evolution. However, learned patterns remain implicit in neural forecasters, making their influence difficult to interpret. Failing to distinguish local evolution can lead to similar forecasts for locations that develop differently, increasing prediction error. This motivates explicit references for distinguishing and interpreting local patterns. Prototypes provide explicit pattern references, but sample- or task-level conditioning leaves open how to adapt to distinct locations and variable groups. To deal with this problem, we design ProtoCast, which uses group-specific prototype banks to adapt local features while accounting for cross-variable dependencies. Specifically, (1) we combine each group's historical features and temporal change with forecast context from cross-variable attention for evolution-aware prototype matching. (2) Sparse mixtures of prototype-specific low-rank operators tailor corrections to each group and location. A gated attention block then exchanges the corrected features across groups before prediction. Each group's correction decomposes into explicit prototype contributions before this fusion. (3) We jointly train the forecasting network and operators from scratch and refit each group's prototypes before every training epoch. Slot tracking preserves prototype-operator associations, making each prototype's contribution to local corrections more interpretable. Experiments on GlobalForecast and TyphoonTracking show improved forecasting accuracy, while prototype deletion links local feature contributions to changes in prediction error. This work advances the development of more accurate and interpretable data-driven weather forecasting models. Project page: https://zhixuanli.github.io/projects/li2026protocast
Nov 19, 2026

Falsifiable Compositional Reasoning in Time Series Forecasting

Time series foundation models as well as the language-guided analytical agents achieve strong accuracy on standard chronological benchmarks. However, models that rank highest on historical validation splits often degrade across non-stationary regime transitions. We hypothesise that this gap arises from using historical backtesting as an open-ended search and model-selection objective instead of using it to test pre-specified compositional assumptions. Real-world time series are generated by recurring latent primitives that co-move within finite historical windows. Under strong historical co-movement, chronological holdout error cannot separate predictors that learn invariant couplings among primitives from predictors that fit transient historical correlations. Searching over large pools of architectures, hyperparameters, or chain-of-thought rules further biases selection toward historical shortcuts. Through theoretical analysis and diagnostic experiments on a controlled multi-primitive benchmark and German day-ahead electricity prices, we show that expanding validation search fails to improve out-of-regime accuracy, and that decomposing primitives without an ex-ante falsification test still breaks down when residual correlations shift. Hence, we propose formulating time series forecasting and reasoning around falsifiable compositional hypotheses that combine neural primitive encoders, explicit coupling constraints, and training-calibrated falsification tests.
Nov 18, 2026

A Verifier from the Past: Test-Time Compute for Time Series Foundation Models

Test-time compute (TTC) improves language models by spending additional inference compute to generate, verify, and select among candidate outputs, but this recipe has not transferred directly to time series forecasting. A language model can verify a proof or vote across samples, but a forecaster must be causal; it cannot use future information to score a candidate. We propose Backtest-Gated Augmentation Averaging (BGAA), a TTC algorithm for time series forecasting that builds a causal verifier from the series's own history, using rolling-origin backtests to score structured views of the context, discard those that underperform the untransformed context, and average the survivors' predictive distributions. BGAA improves every tested TSFM-benchmark combination, reducing MASE by 0.5-1.0% on GIFT-Eval, 1.3-5.0% on fev-bench, and 1.7-5.9% on BOOM, where a 313M model with TTC outperforms the 1B and 2.5B models in its family. We further show that useful views have structured, model-dependent effects consistent with a prior-transport interpretation. On the 53 cross-fittable GIFT-Eval configs, the deployed policy gains only 0.4-1.3% CRPS, versus 1.4-4.1% when choosing the best view for each dataset configuration and 17-19% when choosing the best view separately for each test window. This gap suggests that a major remaining challenge is window-level view selection.
Nov 18, 2026

TRESS: A Fixed-Size History Representation for Irregular Time Series Forecasting

A fixed number of recent observations may span minutes for one variable and days for another. Earlier measurements can still reveal whether the latest value is unusual for that variable. We propose TRESS (Timestamped Recent Events and Statistical Summaries), a forecasting model that retains the latest observations at their original times and summarises the full history of each variable. For a given set of variables and event budget, this representation stays the same size as the record grows. Its query-conditioned event lifting module weights deviations from the latest value according to the age of each observation and the forecast horizon. A channel-specific decoder combines the resulting features with historical statistics and a summary shared across variables to predict a correction to the latest value. Retaining 8 to 16 observations per variable, TRESS records the lowest mean test MSE and MAE among the methods compared on all four benchmarks. Serving 32 successive query times from one encoded history reduces CPU inference time by a factor of 2.12--2.60 across the four datasets. The code is at https://github.com/SSMa913/NICLab-TRESS.
Nov 17, 2026

Which Past Matters? Training History and Multi-Forecast Horizon in Equity Transformers

More historical data can improve estimation while introducing relationships that transfer poorly to current market conditions. We therefore examine how training quantity and historical selection interact with forecast horizon in intraday equity prediction. We deploy twelve configurations which combine four sampling and validation policies with 6, 12 and 24 month training quantities under a fixed cross-sectional Transformer and feature set. A monthly walk-forward experiment produces 264 refits and eight cumulative return forecasts from 5 to 120 minutes. Evaluation covers January 2025 through 20 October 2026 and uses cross-sectional Spearman information coefficients against future five-factor residual returns. Six-month recency sampling leads individual models at 5-15 minutes, while multi recency configurations lead at 20-30 minutes and uniform configurations at 60-120 minutes. Increasing training quantity does not uniformly improve accuracy while an equal-weight ensemble of all twelve models exceeds every individual mean at 15, 20 and 25 minutes, but not at the remaining horizons. These findings support evaluating historical selection jointly with forecast horizon without identifying a universal policy or a causal benefit from recency.
Nov 17, 2026

CorrectCast: Selective Multimodal Context for Scale-Aware Residual Correction in Precipitation Nowcasting

Precipitation nowcasting models suffer from errors in intensity, location, and spatial structure. Addressing these errors by retraining nowcaster backbones can be computationally expensive and requires architecture-specific modifications. To overcome these limitations, we propose CorrectCast, a decoupled residual correction framework that refines completed nowcasts while keeping the base nowcaster frozen. By operating on predicted precipitation fields, CorrectCast requires training only a lightweight corrector, reducing computational overhead while providing a unified correction interface across different nowcasting models. Conditioned on the base nowcast and historical quantitative precipitation estimation (QPE) and precipitable water vapor (PWV) observations, the corrector selects multimodal context and employs scale-aware modulation to estimate residual corrections. On the Mississippi River Basin (MRB) and contiguous United States (CONUS) datasets, CorrectCast improves six base nowcasters spanning extrapolation-based, deterministic, and generative approaches, with absolute gains in mean CSI averaging 0.0233 and 0.0290 across the six models, respectively.
Nov 16, 2026

Benchmarking memory and nonlinear processing capacity of quantum recurrent neural networks for time series prediction

Quantum machine learning has emerged as a promising approach to time series prediction. While quantum reservoir computing has been widely explored for this setting, variational quantum models tailored to sequential data remain comparatively understudied, largely because they are difficult to train. In this work, we present a rigorous benchmark of variational quantum recurrent neural networks (QRNNs) on time series prediction tasks, focusing on two key capabilities: memory and nonlinear processing. We introduce a task-driven methodology that quantifies model capacity, enabling an intuitive and fair comparison across architectures, including classical recurrent baselines. We evaluate the proposed framework on two generic benchmark tasks and on a synthetic dataset describing a chaotic system. Overall, the quantum models achieve performance comparable to their classical counterparts, but classical models scale better with the memory size within the range studied. Notably, one QRNN variant shows higher average capacity on the task designed to jointly probe memory and nonlinear capacity. Moreover, our results show that quantum models offer better capacity when we are able to find a good parameter initialization. However, the increasing sensitivity to initialization leads to a lower probability of finding good solutions as the model size increases. These results provide a principled starting point for designing richer quantum recurrent architectures and for assessing where they may offer practical benefits.
Nov 16, 2026

Wavelet-Energy-Driven Dynamic Patching for Time Series Forecasting

Patch-based transformers tokenise a time series into fixed-size, uniform windows, which commits the model to a single temporal resolution even though real signals are multi-scale. The discrete wavelet transform (DWT) is a natural alternative because it decomposes a signal across scales. A growing body of work makes the wavelet filter learnable, on the assumption that a more flexible wavelet forecasts better. That assumption has rarely been tested in isolation, because a single reported improvement usually credits three things at once. These are the use of any multi-resolution decomposition, the choice of wavelet and training the filter. We introduce a forecaster that turns a learnable-DWT per-scale energy map into variable-length patches by merging adjacent windows that share a dominant scale. Three findings follow. First, most of the improvement is there before any filter is learned. A frozen wavelet beats a fixed-patch baseline on eight of the nine datasets by 1.5% to 7.8%, and on four of them it recovers all of the gain. Traffic is the one exception, where it is 1.2% worse. Second, learning the filter gives a further improvement on only some datasets. It helps on three of the nine datasets and makes Traffic 2.7% worse. Third, having variable-length patches helps, but exactly where their boundaries fall does not. A fixed-size uniform grid, even with its patch size tuned per task, is worse than our variable-length patches on 27 of 36 settings. Recomputing the boundaries from the learned filter at every step, rather than caching them once, never helps. The improvement in wavelet-based forecasting comes from the multi-resolution structure and the variable-length patches it induces, not from making the wavelet learnable. In practice, start from any valid wavelet, cache the patch boundaries once and treat training the filter as a per-dataset decision rather than a default.
Nov 16, 2026

Conformal Expert Pools: Training-Free Probabilistic Time-Series Forecasting Beyond Seasonality

Conformal Seasonal Pools (CSP) showed that a forecaster with no training at all, built only from past values at the same point in the season and from past forecast errors, significantly outperformed DeepNPTS, Amazon's learned non-parametric forecaster. We propose Conformal Expert Pools (CEP), a training-free successor. CEP replaces the single seasonal-naive forecast by a pool of simple forecasters, builds its predictive distribution from their past forecast errors, and adapts both the pool and the errors to the series being forecast. No model is trained and no parameter is learned across series. On the CSP benchmark of 232 datasets and 24,502 forecast windows, CEP is better than CSP on 78% of datasets by CRPS, on 84% by MSIS and on 83% by Winkler score; on a typical dataset its CRPS is 14% lower and its MSIS 23% lower (both p < 1e-22, paired Wilcoxon test), and coverage of 95% intervals rises from 90.1% to 92.2%. CEP has a lower CRPS than DeepNPTS on 88% of datasets, than NPTS on 89% and than SeasonalNPTS on 83%. The improvement persists on dataset families outside the development set and on 36 additional fev-bench files. We also test the published claims that Laplace, an online ensemble forecaster, "beats (almost) everything" and "beats" CSP. On 228 datasets at multi-step horizons the claims do not hold: CSP outperforms Laplace on 60% of datasets by CRPS and on 64% by interval score, and Laplace ranks sixth of seven methods, with the widest intervals of all (69% wider than CEP's). Its point forecast is biased low and on 54% of datasets is less accurate than seasonal naive, and from a cold start it takes about 7,700 times longer than CSP. CEP is better than Laplace on 82% of datasets by CRPS and on 80% by MSIS, with intervals about 40% narrower, and runs 271 times faster.
Nov 16, 2026

FARCH: A Functional ARCH Model for Calibrated Prediction Bands

Forecasting a curve, rather than a handful of scalar summaries of it, requires a notion of uncertainty that is itself functional. We propose FARCH, a functional nonlinear autoregressive model with conditional heteroscedasticity, in which the conditional mean and the conditional volatility of the next curve are forecasted jointly and nonparametrically, by reusing a single kernel weighting scheme --- built on a wavelet semi-metric between curves --- first on the curves themselves and then on their squared residuals. From this pair of estimators we derive two constructions of prediction bands: a pointwise band, obtained from marginal scenario quantiles, and a uniform, sup-norm band, calibrated for decisions that concern the entire forecasted path rather than any single instant. Applied to day-ahead forecasting of half-hourly French national electricity demand (1996--2009), and compared, under an identical protocol, against four existing prediction-interval constructions that do not model the conditional variance explicitly, the uniform FARCH band attains pointwise coverage statistically indistinguishable from its nominal target at the 80%, 90% and 95% levels, and a curve-wise coverage of 90.1%, significantly above the 70.8% reached by the best competing method, at the cost of a wider interval. Pointwise bands, in contrast, undercover regardless of whether the variance is modelled explicitly, indicating that state-dependent volatility and simultaneous calibration are two distinct ingredients that a trustworthy prediction band must combine.
Nov 16, 2026

MSAR: Next-Scale Autoregressive Forecasting for Time Series via Modular Multi-Scale Decoupling

Time series forecasting underpins critical applications in finance, energy, healthcare, and transportation. Although deep models have achieved strong results, most adopt single-scale modeling or restrict multiscale processing to the input side, causing a misalignment between multiscale inputs and single-scale outputs and limiting predictive power. We introduce the Modular Scale-wise Autoregressive Framework (MSAR), a model-agnostic design that forecasts progressively across multiple temporal resolutions. MSAR offers three advantages: (1) scale-wise aligned modeling, which disentangles heterogeneous temporal patterns by aligning inputs and outputs at each scale; (2) scale-wise autoregression, where coarse-scale predictions guide finer-scale forecasting through hierarchical information flow; and (3) a modular architecture, enabling seamless integration with diverse backbones such as CNNs, MLPs, and Transformers. Extensive experiments across a broad set of datasets and forecasting models demonstrate that MSAR achieves consistent improvements in both accuracy and inference efficiency, validating the effectiveness of scale-aligned autoregression for multiscale time series forecasting. The code is available at: https://github.com/Oryxo/MSAR.
Nov 16, 2026

A Temporal Convolutional Network with Multi-Head Attention Model: A Surrogate for Parallel DEVS Semiconductor Fab Simulation

Semiconductor factories process many kinds of wafers organized into lots of various sizes and configurations. Different kinds of machines, each with its own operations, process wafers for hours to days, and the wafers undergo many incremental fabrication steps (with re-entry) and flows spanning days to weeks across connected machines. Fast and accurate predictions of throughput (THTH) and turnaround time (TATA) dynamics for individual machines and their combinations are essential for factory operations. To use machine learning methods for this purpose, this paper presents a Temporal Convolutional Network with Multi-head Attention (TCNA) architecture. It uses simulated datasets from a benchmark Parallel DEVS (PDEVS) semiconductor fabrication factory model. The factory processes configurable wafer lots with uniform and sinusoidal wafer arrival profiles and may undergo preventive maintenance. The wafer lot size, arrival profile, and factory maintenance policy are used as static covariates. TCNA is applied to a selected set of simulated discrete-event throughput and turnaround time trajectories. The THTH and TATA time series predicted by TCNA, evaluated with the common MAPEMAPE, MSEMSE, and R2R^2 measures, show comparable or higher accuracy relative to other model architectures, with similar parameter counts and epoch times; the TCNA model size and execution time are similar to those of other architectures such as the Temporal Fusion Transformer. Although time series predictions are less accurate than high-resolution simulations, they can be deployed at a much lower computational cost. The TCNA achieves orders-of-magnitude speedup over PDEVS simulation with an acceptable loss of accuracy, enabling frequent product scheduling, capacity planning, and maintenance planning.
Nov 14, 2026

A Hybrid State Space-Attention Architecture for Forecasting the Trajectories of Battery Degradation

Early-life battery degradation trajectory forecasting aims to predict a battery cell's full state-of-health (SOH) trajectory from its initial charge-discharge cycles, potentially reducing cell qualification and evaluation time from months to days. Despite recent advances, accurate forecasting remains challenging for emerging battery chemistries, data-scarce domains, and cells exhibiting degradation behaviors beyond those observed during training. Existing state-of-the-art approaches increasingly rely on model scaling, large language model-derived aging-condition embeddings, and extensive fold-specific hyperparameter tuning, raising concerns about computational efficiency, reproducibility, and generalization. This work introduces DecayFormer, a compact hybrid state-space and attention architecture designed for early-life degradation trajectory forecasting. DecayFormer combines cycle-level selective recurrence, using a single scalar state per channel, with an attention mechanism queried by the recurrent state. It further incorporates state-of-charge (SOC) interval statistics and a master curve head that jointly estimates a scalar lifetime and a monotonically non-increasing SOH profile parameterized by normalized lifetime. We evaluate DecayFormer on the BatteryLife benchmark using all 22 aging-condition-exclusive folds, with a single run per fold and one fixed configuration across all folds, without fold-specific tuning. DecayFormer achieves mean test MAPE values of 2.25% on Li-ion, 2.76% on CALB, 0.95% on Na-ion, and 7.19% on Zn-ion. These results demonstrate the potential of compact sequence models to deliver accurate early-life degradation forecasts across diverse battery chemistries and aging conditions, without relying on extensive model scaling or per-fold optimization.
Nov 13, 2026

SWIFT: Sample-Wise Instance Fine-Tuning of Time-Series Foundation Models for Forecasting

While current few-shot adaptation strategies for time series forecasting foundation models are predominantly dataset-centric, the necessity of obtaining an additional dataset for the finetuning process poses a significant hurdle for adoption. To address this constraint, we propose Sample-Wise Instance Fine-Tuning (SWIFT), a novel forecasting adaptation framework that achieves fine-grained specialization by leveraging structural priors from the target instance's own history. This shift allows the fine-tuning process to be entirely transparent to the end-user, which significantly enhances the ease of adoption. To implement this framework, SWIFT synergizes Self-Retrieval for historical mining and Frequency-Aware Loss for spectral consistency, while employing Deviation-Aware Dynamic Fusion to balance adaptation with general knowledge preservation. Extensive experiments across four TSFM backbones and three long-history benchmark subsets validate SWIFT, yielding average MASE and WQL reductions of 6.41% and 4.83% across 24 combinations of four backbones and six Chronos Benchmark II datasets. To the best of our knowledge, SWIFT is the first framework to investigate sample-wise instance fine-tuning of time-series foundation models, specializing the model for an individual prediction window using only the corresponding forecasting sample's own observed history.
Nov 13, 2026

FlexCast: Flexible Context Learning for Time Series Foundation Models

Time series foundation models (TSFMs) based on autoregressive (AR) next-token prediction entangle two factors: variable-context conditioning and prediction-unit factorization with autoregressive rollout. Strong AR performance therefore does not reveal whether generalization stems from the flexible training distribution, the local prediction target, the left-to-right inference process, or their combination. We propose FlexCast, a flexible context-target learning framework that disentangles these factors. FlexCast reformulates forecasting as variable-length masked prediction over time series patches: during pretraining, it dynamically samples visible context and masked target patches, exposing the model to diverse context-target configurations without sequential decoding. At inference time, the same pretrained model supports one-step forecasting, confidence-based iterative refinement, and next-token-style generation through a unified mask-and-predict interface, enabling controlled comparison of inference schedules under identical architecture, data, and objective. Evaluated on BLAST and multiple zero-shot benchmarks, FlexCast achieves accuracy competitive with or better than strong TSFM baselines while substantially reducing inference latency. Ablations confirm that context diversity and contiguous target masking drive performance, while left-to-right decoding does not yield consistent gains once flexible context-target relationships have been learned, suggesting that the key benefit of AR-style TSFMs lies in flexible training rather than sequential inference. The code is available at https://github.com/BenchCouncil/FlexCast.
Nov 12, 2026

Learning Temporal Dynamics with Derivative Representations for Time Series Forecasting

Time series forecasting aims to predict the future evolution of complex systems by learning their underlying dynamics from past observations. Despite recent advances in deep learning, Transformers in particular still struggle to capture the temporal evolution of system dynamics. Existing approaches have been largely driven by architectural innovation, with comparatively little attention paid to training strategies that explicitly exploit the temporal structure governing system dynamics. In this work, we introduce DeltaTSF, a dynamics-aware forecasting framework based on discrete derivative representations of time series. Specifically, DeltaTSF jointly learns the series and their discrete temporal derivatives across both input and forecast horizons within a unified, model-agnostic framework. By coupling these representations over time, the model is encouraged to maintain temporal consistency, a property we show to induce a form of discrete Taylor consistency during training. This formulation enables the model to better reflect the underlying system dynamics, resulting in more stable and physically coherent forecasts. Notably, DeltaTSF does not require explicit derivative computation and introduces no additional computational overhead at inference time. Extensive experiments across diverse benchmarks and backbone architectures demonstrate that DeltaTSF consistently improves forecasting performance and yields more dynamically coherent predictions than strong SOTA methods. These results highlight the effectiveness of enforcing dynamic coherence through derivative representations as a model-agnostic paradigm that operates alongside and complements architectural design. Source code will be made publicly available upon acceptance.
Nov 12, 2026

Climate Temporal Temperature Prediction via an Adaptive Stacked Bidirectional Residual Mamba Network

Accurate temperatures forecasting relies on traditional meteorological parameters that are essential for monitoring weather informatics and guiding forecasting efforts. This study presents Deep DualMamba, a novel deep learning framework for temporal temperature prediction, built upon an enhanced variant of the recently proposed Mamba network. Unlike conventional Transformer-based architectures, which dominate foundation models but suffer from computational inefficiencies on long sequences, Deep DualMamba leverages selective structured state space models (SSMs) to achieve both scalability and performance. We identify a core limitation of prior subquadratic-time models-namely, their inability to perform long-term and robust forecasting and introduce two key innovations. First, we propose an adaptive bidirectional Mamba as encoders to learn complex and efficient features. Second, we develop a hardware-aware, parallel recurrent algorithm to compensate for the loss of convolutional efficiency, ensuring practical viability at scale. Last, the Deep DualMamba's approximation ability and convergence is theoretical analyzed. These design choices culminate in a simplified, attention- and MLP-free neural architecture that exhibits linear complexity with sequence length and significantly improved inference throughput. To demonstrate its applicability in climate informatics, we implement a full forecasting pipeline using historical meteorological data from UK monitoring stations.
Nov 11, 2026

MORA: Correcting Time Series Forecasts with Coarse Future Signals

Recent architectural advances have improved time-series forecasting. However, our analysis of forecasters with diverse architectures reveals a common pattern: a substantial share of residual energy is concentrated in a small number of low-frequency components. Simple linear probes further show that the forecasters' historical inputs still contain predictive information about errors in the forecast mean. Building on these findings, we propose Model-Output Residual Adaptation (MORA), a model-agnostic residual adapter that refines the outputs of frozen forecasters without requiring additional inputs. MORA estimates segment-wise means of the target sequence through structural extrapolation and learned estimation, rather than directly predicting backbone-specific residuals. It then converts discrepancies between these estimates and the corresponding statistics of the backbone forecast into stagewise corrections, scaling each update according to estimation reliability. Extensive experiments on twelve benchmarks with six forecasting backbones show average MSE reductions of 7.2% for long-term forecasting and 28.1% for short-term forecasting, with lower MSE in 96.2% of the 288 evaluated cases. The benefits also extend to three time-series foundation models evaluated at different context lengths. Together, these results demonstrate that predictable coarse-scale residuals can be effectively exploited to improve existing forecasts. Code repository: https://github.com/hugerli-1/MORA.
Nov 11, 2026

Quantum Recurrent Architectures for Financial Time Series Prediction

This study evaluates quantum classical hybrid recurrent architectures for financial time-series forecasting: a Quantum Long Short-Term Memory (QLSTM) network and Quantum Reservoir Computing (QRC) with lag-based inputs. Classical lag vectors are encoded into quantum states via amplitude encoding, which compactly represents normalized observations but requires nontrivial state preparation; the QLSTM gates and the QRC reservoir are implemented with parameterized quantum circuits, while classical optimizers train the interface and readout layers. On a monthly revenue series extended to 1056 months by a Gaussian-process--hidden-Markov generator, with a fixed chronological 70/15/15 split and one-step-ahead forecasts, we compare both models against a full-size LSTM, a classical control matched in parameter budget and interface architecture, a small plain LSTM, and classical reservoir computing, over ten seeds and against persistence, seasonal-naive and SARIMA benchmark forecasts. The QLSTM, with 2,105 parameters, is statistically indistinguishable from an LSTM 126 times larger, but equally indistinguishable from the matched classical control and from a plain LSTM of the same budget, so its parameter efficiency cannot be attributed to the quantum circuits. The quantum reservoir is worse than a classical reservoir of equal feature dimension in all ten seeds and worse than SARIMA, whereas all recurrent models and both classical reservoirs improve on SARIMA by about 2% in RMSE. In a 24-month rolling-origin backtest the classical LSTMs degrade more slowly than the QLSTM. We find no evidence of a quantum accuracy or parameter-efficiency advantage in this low-dimensional setting.
Nov 11, 2026

SkinAdvisor: An AI-Driven Decision Framework for Skin Production Planning via Multi-Modal Transformer Sales Forecasting

Skins, cosmetic in-game items, are a primary revenue stream for modern online games, yet production decisions (which skins to make, when to release, how to price) remain driven by intuition and retrospective analytics. We present SkinAdvisor, an AI-driven decision framework for skin production planning, powered by SkinCast, a deployed multi-modal Transformer forecasting 180-day per-day sales trajectories for unreleased skins. SkinCast couples four encoders: a pretrained Informer for player-preference time series, a pretrained Player-Hero-Skin graph embedder for relational priors, and two lightweight Transformers for skin attributes and release context, fused through a shared backbone with a local-plus-global day-x positional encoding and decoded by an N-layer Informer in a single non-autoregressive pass. SkinAdvisor plugs these forecasts into a serving pipeline (Select -> Fill -> Infer -> Sales -> ROI -> Ranking) that ranks candidate skins by forecast-based ROI and supports multi-candidate comparison. The system has been in production for a full release cycle. Offline experiments show SkinCast reduces RMSE by ~20% over the strongest baseline (Informer) and ~60% over classical methods; stress tests on a 16-core CPU server confirm p99 latency under 8ms and peak throughput above 1100 rps. We report two deployment lessons: order-of-magnitude accuracy suffices for production decisions, and pretrained embeddings can leak temporal information, requiring season-based splitting to prevent.
Nov 10, 2026

Generalize Then Personalize: Wearable-Based Heart Rate Forecasting Under Scarce Ruck-March Data

Heart rate (HR) forecasting may provide advance information about near-future cardiovascular response, potentially supporting physiological monitoring and workload-management decisions during ruck marching. However, robust model development is constrained by the difficulty of collecting large-scale ruck-march datasets, creating a scarce-data problem in which models must generalize across participants while adapting to individuals from limited personal data. To address this problem, we benchmark population-only training (POT), personalized-from-scratch training (PFST), and population-pretrained fine-tuning (PPFT) for HR forecasting during loaded ruck marching using data from eight active-duty U.S. Army medics across thirteen sessions. LSTM, Attention-LSTM, and CNN-LSTM backbones were evaluated against a persistence baseline at 30 s and 60 s horizons. Across all neural architectures and horizons, PPFT yielded lower mean MAE, RMSE, and endpoint MAE than both POT and PFST. Participant-level paired analysis after averaging across three training seeds showed that, relative to PFST, PPFT reduced trajectory MAE by 2.22-4.30 bpm and endpoint MAE by 1.24-2.88 bpm. All six trajectory-MAE comparisons remained significant after Holm adjustment (padjp_{adj} = 0.0469), while four of six endpoint-MAE comparisons remained significant after adjustment. At 30 s, persistence achieved the lowest MAE (3.555 ±\pm 1.274 bpm), while CNN-LSTM PPFT achieved the lowest RMSE (4.669 ±\pm 1.916 bpm); LSTM PPFT achieved the lowest endpoint MAE (5.093 ±\pm 2.365 bpm). At 60 s, Attention-LSTM PPFT achieved the lowest MAE (5.573 ±\pm 1.939 bpm) and RMSE (7.134 ±\pm 2.260 bpm), while CNN-LSTM PPFT achieved the lowest endpoint MAE (7.553 ±\pm 2.357 bpm). These results suggest that population pretraining followed by participant-specific fine-tuning may provide a practical strategy for scarce-data wearable HR forecasting.
Nov 10, 2026

TimeSeek: Multi-Task and Multi-Domain Pretraining for Generalized Time Series Analytics

Time series analytics has been widely applied across various domains. While existing time series foundation models (TSFMs) have demonstrated promising results, they still face two major challenges: 1) Most TSFMs mainly rely on stacking Transformer layers to increase parameter size, lacking the ability to adaptively model complex time series patterns from multi-domain data. 2) Current TSFMs are primarily pre-trained for specific tasks, e.g., forecasting or anomaly detection, and a few models support multiple tasks, but they fail to fully leverage task specificity and commonality, resulting in suboptimal performance. To address these challenges, we propose TimeSeek, a novel generalized time series analytics model that adaptively models time series patterns from multi-domain data and unifies multiple tasks in pre-training. Conceptually, TimeSeek is guided by selective information sharing under domain and task heterogeneity. From the perspective of multi-domain data, we propose a Multi-Domain Feature Extractor that adaptively captures temporal dependencies and channel correlations based on domain-specific characteristics, including progressive multi-scale modeling and adaptive slot channel modeling. From the perspective of the diversity of tasks, we propose multi-task collaborative training, which leverages task specificity and task commonality through task customization and cross-task training, effectively unifying multiple tasks in pre-training and better adapting to diverse downstream scenarios. After pre-training on large-scale time series data, TimeSeek achieves state-of-the-art performance in forecasting, anomaly detection, and classification tasks across both few-shot learning and zero-shot situations, together with favorable inference efficiency.
Nov 10, 2026

Quantifying the Sampling Frequency Invariance of Time-Series Foundation Models for Sensor Data Applications

Time-Series Foundation Models (TSFMs) are increasingly used in sensor applications, where sampling frequency (SF) varies across sensors, devices, and operating modes (e.g., power saving). Ideally, TSFMs must remain effective during such SF changes, provided the SF retains task-critical bandwidth. In this paper, we stress-test 4 prominent TSFMs: MOMENT, Chronos-2, Moirai-MoE, and TimesFM, to quantify their SF-invariance and investigate whether their behavior tracks the resulting loss of information bandwidth rather than the SF value itself. Our stress-test probes 3 dimensions: (i) the shift in embeddings, (ii) the degradation in forecasting performance, and (iii) the degradation in downstream performance, quantified by three novel metrics: FSI, FDI, and CDI, respectively. To our knowledge, this is the first study to place SF-invariance at the core of TSFM evaluations. Our results characterize: (i) how SF-invariance depends on the margin between signal's native-SF and the bandwidth its task requires, (ii) how much of that invariance the choice of TSFM explains, and (iii) how much of TSFM's performance degradation under a lower SF reflects information loss rather than a limitation of their abilities. These findings reframe SF-invariance as a measurable, signal-aware property of the (signal X TSFM) pair, and guide practitioners on when TSFMs display SF-invariance.
Nov 9, 2026

Continuous Horizon Forecasting

Multi-horizon forecasting has been dominated by discrete forecast strategies that treat model outputs as fixed-vectors or recurrently create them. In this paper, we introduce the Continuous Horizon (CH) strategy that generalizes upon recursive, direct, and multi-step strategies by redefining forecasts as a continuous operator and the forecast horizon as a query variable rather than an index. We present the theoretical foundations for the CH strategy, a Universal Operator Approximation Theorem, and a continuous-horizon bias-variance decomposition that formally proves the estimation variance improvements of CH over the discrete strategies. We validate the CH strategy empirically through a simulation study, and on 16 large-scale datasets from the M-series benchmark, showing how CH achieves competitive RMSE accuracy, while uniquely enabling parallel inference, targeted horizon querying, and interpolation and extrapolation across unseen horizons.
Nov 9, 2026

Do Time-Series Foundation Models Forecast Index-Futures Returns? A Preregistered Evaluation of Models, Metrics, and Input Representations

We ask whether the apparent forecasting skill of frozen time-series foundation models (TSFMs) on intraday index-futures returns is incremental skill. In a pre-registered evaluation on E-mini Nasdaq-100 (NQ) and E-mini S&P 500 (ES) futures, a common panel of five zero-shot models (Chronos-Bolt, Chronos-2, TimesFM 2.5, Moirai 2.0 and TiRex), plus separate experiments with Kronos, is compared with a no-change forecast at 5-120 minutes, using native prices, log returns, volatility-standardised returns and a volume clock as inputs. On a blind set opened once, 0 of 160 model-representation-horizon-market cells beat no-change, and a Hansen SPA test gives p = 1.000 on both markets; a later retrospective hold-out, scored as a descriptive check, gives 0 of 120. Input representations change forecast magnitude but not accuracy, and a transformation-aware ridge regression does as well. Several standard measures nevertheless suggest skill. Price-level correlation exceeds 0.9996, but no-change scores higher. Within-day RankIC across forecast origins rewards a trivial revert-to-mean rule (+0.35 at 30 minutes); we derive the expected within-day covariance under a random walk in closed form, and a sign-randomised null reproduces about 95% of the observed value. One model beats HAR for 5-minute volatility, but not the EWMA its input already contains. On strongly planted synthetic signals the models recover part of the signal at 5 minutes, and recovery deteriorates at 30 minutes. Benchmark-level success of TSFMs should therefore not be assumed to carry over to low-signal domains such as intraday returns.
Nov 8, 2026

Means vs. Foundation Models for Emergency Call Forecasting

This study investigates whether generalized means can achieve forecasting performance comparable to that of TimeGPT for daily emergency call volume prediction. Specifically, two families of generalized means, the Hölder means (HM) and the Lehmer means (LM), are evaluated against fine-tuned TimeGPT across forecasting horizons ranging from one to seven days. Forecasting performance is assessed on a real-world emergency call dataset using multiple error metrics, and Diebold-Mariano test. Experimental results show that HM and LM achieve nearly identical predictive performance across all forecasting horizons. For short-term forecasting, both generalized means outperform TimeGPT, with reductions of mean absolute error (MAE) approximately 1.24% and 2.48% at horizons of one and three days, respectively. Conversely, TimeGPT achieves a lower error at the seven-day horizon, with an descresing of MAE by 1.98%. However, Diebold-Mariano tests reveal no statistically significant differences between the methods at the evaluated horizons. These findings suggest that generalized means can provide competitive forecasting performance compared with a energyvore foundation model.
Nov 8, 2026

Daily Emergency Call Volume Forecasting using Univariate and Multivariate TimeGPT

This study investigates the potential of TimeGPT, a pretrained foundation model for timeseries, for forecasting daily emergency call volumes. Real-world emergency call data, comprising 2,101 daily observations, are used to evaluate both univariate and multivariate forecasting settings under one-step-ahead and seven-day multi-step-ahead scenarios. The multivariate forecasting incorporates temporal, meteorological, demographic, and socioeconomic variables. A chronological training, validation, and test split is adopted to prevent information leakage, while hyperparameters are selected using the validation set through grid search and early stopping when applicable. Forecasting performance is evaluated using multiple error metrics. Its forecasting performance is compared with several established forecasting approaches, including Seasonal Autoregressive Integrated Moving Average with Exogenous Regressors (SARIMAX), Poisson regression (PR), Negative Binomial regression (NBR), Extreme Gradient Boosting (XGBoost), Multilayer Perceptron (MLP), and Long Short-Term Memory (LSTM). The results show that SARIMAX achieves the best overall performance in the univariate setting. In the multivariate setting, NBR and PR provide best performance. These findings indicate that the large-scale pretraining of TimeGPT do not necessarily translate into superior forecasting accuracy for emergency call volumes. Frugal methods trained solely on the available data remain better for this application.
Nov 8, 2026

ClusLoad: Cluster-Guided Model Sharing for Scalable Distribution Transformer Load Forecasting

Large-scale device-level load forecasting for distribution transformers is challenged by substantial cross-device heterogeneity and the deployment burden of maintaining device-specific forecasting models. Existing approaches typically rely either on device-specific models, which are difficult to scale to large transformer populations, or on global models, which reduce deployment complexity but may overlook heterogeneous load patterns across devices. To address this trade-off, this study develops a trend-clustered forecasting framework that enables model sharing among transformers with similar temporal behaviors. Representation Learning for Time Series using Time-Embeddings (T-Rep) is employed to learn compact temporal representations of transformer load series, which are subsequently grouped using a Gaussian mixture model (GMM). Representative transformers are sampled within each cluster to train cluster-specific Patch Time Series Transformer (PatchTST) models while retaining device-level predictions. The framework is evaluated using real-world 15-min operational data from nearly one thousand distribution transformers. The proposed framework reduces mean absolute error (MAE) by approximately 20% relative to the direct PatchTST baseline, achieves a high-load warning accuracy of 95.66%, while requiring only seven cluster-specific forecasting models. These results demonstrate that behavior-based device organization provides an effective balance between forecasting accuracy, deployment scalability, and risk-oriented operational support for distribution transformer management.
Nov 8, 2026

StatPrompt-LLM: A Lightweight Statistical Refinement for Foundation-Style Time-Series Forecasting

Time-series forecasting has recently seen a paradigm shift with the introduction of foundation-style models that adapt large pretrained language architectures to sequential data. However, effectively exposing fundamental temporal structures, such as trend, scale, and temporal dependence, to these models remains an open challenge, particularly when the available observations are limited. In this paper, we conduct an exploratory empirical study across different generations of forecasting architectures and propose StatPrompt-LLM, a lightweight statistical refinement of Time-LLM. StatPrompt-LLM augments the existing Time-LLM prompting mechanism with dataset-level statistical information and introduces a zero-initialised autoregressive bypass as a complementary linear forecasting pathway. We evaluate our framework on electricity and meteorological time-series benchmarks, including ETTh1, ETTh2, and Weather, under full-prefix and reduced-prefix configurations. Under the 10% reduced-prefix configuration, StatPrompt-LLM achieved lower in-sample prediction errors than Time-LLM, with unweighted average relative reductions across datasets of 40.2% in MSE and 21.4% in MAE. Full-prefix results were more dataset-dependent. Ablation analyses revealed relatively modest and dataset-dependent changes when selected additional statistical descriptor groups were masked within the combined architecture. Computational profiling indicated additional model parameters and runtime overhead relative to Time-LLM.