arXiv subjects

Willa Potosnak

Publications and source records attributed to Willa Potosnak.

12 recordsLinked to original sources

arXiv2026

Scale-Invariant Training for Time Series Foundation Models

Time series foundation models (TSFMs) are trained on large collections of time series datasets that span various morphologies and domains. This setting exposes models to series whose scales -- typical magnitudes of their values -- can differ substantially. Affine scaling methods such as Reversible Instance Normalization (ReVIN) scale model inputs and reverse the transform before computing the loss. We show that this inversion multiplies each series' gradient by $b^p$ relative to loss on scaled targets, where $b$ is the scaling denominator (e.g., standard deviation) and $p$ is the loss degree. We call this scale-contaminated training (ScaleCon), because the scale of each series consequently becomes an importance weight, causing high-scale series to dominate training. For any scale-equivariant scaler and residual loss that is homogeneous of degree $p$, including MSE, MAE, and Quantile Loss, we prove that computing loss on scaled targets makes every mini-batch gradient and, consequently, the full optimization trajectory invariant to arbitrary independent rescaling of the training series, yielding scale-invariant training (ScaleIn). Notably, existing TSFMs use both objectives, with neither consistent reporting nor a common convention on how to compute training loss. We isolate the convergence disparity induced by ScaleCon and its correction under ScaleIn in controlled studies on synthetic and real data. In pretraining across four TSFM architectures, ScaleIn lowers MASE in all 24 architecture-benchmark comparisons, with average reductions across TSFMs of 18.8% on GIFT-Eval and 21.9% on the M-competitions. The gains extend to supervised neural forecasting, where it lowers MASE in 16 of 20 matched settings. Most existing time series forecasting pipelines can adopt ScaleIn with a one-line code change.

cs.LG↗

arXiv2026

Forking-Sequences: Statistically and Computationally Efficient Multi-Horizon Forecasting with Reduced Volatility

While accuracy is a critical requirement for time series forecasting, an equally important desideratum is reasonable forecast volatility across forecast creation dates (FCDs). Even highly accurate models can produce erratic revisions between FCDs, undermining trust and disrupting downstream decision-making. To improve the volatility of forecast revisions, state-of-the-art models like MQCNN, MQT, and SPADE employ a powerful yet underexplored neural network architectural design: forking-sequences. This architectural design jointly encodes and decodes the entire time series across all FCDs, producing an entire multi-horizon forecast grid in a single forward pass. This approach contrasts with conventional neural forecasting methods that process FCDs independently, generating only a single multi-horizon forecast per forward pass. In this work, we formalize the forking-sequences design and motivate its broader adoption by introducing a metric for quantifying excess volatility in forecast revisions and by providing theoretical and empirical analysis. We theoretically motivate three key benefits of forking-sequences: (i) reduced forecast volatility through ensembling; (ii) gradient variance reduction, improving the statistical efficiency of the training procedure; and (iii) improved inference computational efficiency. We validate the benefits of forking-sequences compared to baseline window-sampling on the M-series benchmark, using 16 datasets from the M1, M3, M4, and Tourism competitions. We observe median sCRPS improvements across datasets of 46.2%, 49.3%, 28.6%, 24.7%, and 6.4% for RNN, LSTM, CNN, Transformer, and State Space-based architectures, respectively. We then show that forecast ensembling during inference can reduce median forecast volatility by 13.2%, 13.0%, 10.9%, 10.2%, and 11.2% for these respective models trained with forking-sequences, while maintaining accuracy.

cs.LG↗

arXiv2026

MICA: Multivariate Infini Compressive Attention for Time Series Forecasting

Multivariate forecasting with Transformers faces a core scalability challenge: modeling cross-channel dependencies via attention compounds attention's quadratic sequence complexity with quadratic channel scaling, making full cross-channel attention impractical for high-dimensional time series. We propose Multivariate Infini Compressive Attention (MICA), an architectural design to extend channel-independent Transformers to channel-dependent forecasting. By adapting efficient attention techniques from the sequence dimension to the channel dimension, MICA adds a cross-channel attention mechanism to channel-independent backbones that scales linearly with channel count and context length. We evaluate channel-independent Transformer architectures with and without MICA across multiple forecasting benchmarks. MICA reduces forecast error over its channel-independent counterparts by 5.4% on average and up to 25.4% on individual datasets, highlighting the importance of explicit cross-channel modeling. Moreover, models with MICA rank first among deep multivariate Transformer and MLP baselines. MICA models also scale more efficiently with respect to both channel count and context length than Transformer baselines that compute attention across both the temporal and channel dimensions, establishing compressive attention as a practical solution for scalable multivariate forecasting.

cs.LG↗

arXiv2025

Global Deep Forecasting with Patient-Specific Pharmacokinetics

Forecasting healthcare time series data is vital for early detection of adverse outcomes and patient monitoring. However, it can be challenging in practice due to variable medication administration and unique pharmacokinetic (PK) properties of each patient. To address these challenges, we propose a novel hybrid global-local architecture and a PK encoder that informs deep learning models of patient-specific treatment effects. We showcase the efficacy of our approach in achieving significant accuracy gains in a blood glucose forecasting task using both realistically simulated and real-world data. Our PK encoder surpasses baselines by up to 16.4% on simulated data and 4.9% on real-world data for individual patients during critical events of severely high and low glucose levels. Furthermore, our proposed hybrid global-local architecture outperforms patient-specific PK models by 15.8%, on average.

cs.LG↗

arXiv2025

A More Realistic Evaluation of Cross-Frequency Transfer Learning and Foundation Forecasting Models

Cross-frequency transfer learning (CFTL) has emerged as a popular framework for curating large-scale time series datasets to pre-train foundation forecasting models (FFMs). Although CFTL has shown promise, current benchmarking practices fall short of accurately assessing its performance. This shortcoming stems from many factors: an over-reliance on small-scale evaluation datasets; inadequate treatment of sample size when computing summary statistics; reporting of suboptimal statistical models; and failing to account for non-negligible risks of overlap between pre-training and test datasets. To address these limitations, we introduce a unified reimplementation of widely-adopted neural forecasting networks, adapting them for the CFTL setup; we pre-train only on proprietary and synthetic data, being careful to prevent test leakage; and we evaluate on 15 large, diverse public forecast competition datasets. Our empirical analysis reveals that statistical models' accuracy is frequently underreported. Notably, we confirm that statistical models and their ensembles consistently outperform existing FFMs by more than 8.2% in sCRPS, and by more than 20% MASE, across datasets. However, we also find that synthetic dataset pre-training does improve the accuracy of a FFM by 7% percent.

cs.LG↗

arXiv2025

Investigating Compositional Reasoning in Time Series Foundation Models

Large pre-trained time series foundation models (TSFMs) have demonstrated promising zero-shot performance across a wide range of domains. However, a question remains: Do TSFMs succeed by memorizing patterns in training data, or do they possess the ability to reason about such patterns? While reasoning is a topic of great interest in the study of Large Language Models (LLMs), it is undefined and largely unexplored in the context of TSFMs. In this work, inspired by language modeling literature, we formally define compositional reasoning in forecasting and distinguish it from in-distribution generalization. We evaluate the reasoning and generalization capabilities of 16 popular deep learning forecasting models on multiple synthetic and real-world datasets. Additionally, through controlled studies, we systematically examine which design choices in 7 popular open-source TSFMs contribute to improved reasoning capabilities. Our study yields key insights into the impact of TSFM architecture design on compositional reasoning and generalization. We find that patch-based Transformers have the best reasoning performance, closely followed by residualized MLP-based architectures, which are 97\% less computationally complex in terms of FLOPs and 86\% smaller in terms of the number of trainable parameters. Interestingly, in some zero-shot out-of-distribution scenarios, these models can outperform moving average and exponential smoothing statistical baselines trained on in-distribution data. Only a few design choices, such as the tokenization method, had a significant (negative) impact on Transformer model performance.

cs.LG↗

arXiv2025

Exploring Representations and Interventions in Time Series Foundation Models

Time series foundation models (TSFMs) promise to be powerful tools for a wide range of applications. However, their internal representations and learned concepts are still not well understood. In this study, we investigate the structure and redundancy of representations across various TSFMs, examining the self-similarity of model layers within and across different model sizes. This analysis reveals block-like redundancy in the representations, which can be utilized for informed pruning to improve inference speed and efficiency. Additionally, we explore the concepts learned by these models - such as periodicity and trends - and how these can be manipulated through latent space steering to influence model behavior. Our experiments show that steering interventions can introduce new features, e.g., adding periodicity or trends to signals that initially lacked them. These findings underscore the value of representational analysis for optimizing models and demonstrate how conceptual steering offers new possibilities for more controlled and efficient time series analysis with TSFMs.

cs.LG↗

arXiv2024

Implicit Reasoning in Deep Time Series Forecasting

Recently, time series foundation models have shown promising zero-shot forecasting performance on time series from a wide range of domains. However, it remains unclear whether their success stems from a true understanding of temporal dynamics or simply from memorizing the training data. While implicit reasoning in language models has been studied, similar evaluations for time series models have been largely unexplored. This work takes an initial step toward assessing the reasoning abilities of deep time series forecasting models. We find that certain linear, MLP-based, and patch-based Transformer models generalize effectively in systematically orchestrated out-of-distribution scenarios, suggesting underexplored reasoning capabilities beyond simple pattern memorization.

cs.LG↗

arXiv2024

Towards Long-Context Time Series Foundation Models

Time series foundation models have shown impressive performance on a variety of tasks, across a wide range of domains, even in zero-shot settings. However, most of these models are designed to handle short univariate time series as an input. This limits their practical use, especially in domains such as healthcare with copious amounts of long and multivariate data with strong temporal and intra-variate dependencies. Our study bridges this gap by cataloging and systematically comparing various context expansion techniques from both language and time series domains, and introducing a novel compressive memory mechanism to allow encoder-only TSFMs to effectively model intra-variate dependencies. We demonstrate the benefits of our approach by imbuing MOMENT, a recent family of multi-task time series foundation models, with the multivariate context.

cs.LG↗

arXiv2022

auton-survival: an Open-Source Package for Regression, Counterfactual Estimation, Evaluation and Phenotyping with Censored Time-to-Event Data

Applications of machine learning in healthcare often require working with time-to-event prediction tasks including prognostication of an adverse event, re-hospitalization or death. Such outcomes are typically subject to censoring due to loss of follow up. Standard machine learning methods cannot be applied in a straightforward manner to datasets with censored outcomes. In this paper, we present auton-survival, an open-source repository of tools to streamline working with censored time-to-event or survival data. auton-survival includes tools for survival regression, adjustment in the presence of domain shift, counterfactual estimation, phenotyping for risk stratification, evaluation, as well as estimation of treatment effects. Through real world case studies employing a large subset of the SEER oncology incidence data, we demonstrate the ability of auton-survival to rapidly support data scientists in answering complex health and epidemiological questions.

cs.LG↗

arXiv2021

The World as a Graph: Improving El Niño Forecasts with Graph Neural Networks

Deep learning-based models have recently outperformed state-of-the-art seasonal forecasting models, such as for predicting El Niño-Southern Oscillation (ENSO). However, current deep learning models are based on convolutional neural networks which are difficult to interpret and can fail to model large-scale atmospheric patterns. In comparison, graph neural networks (GNNs) are capable of modeling large-scale spatial dependencies and are more interpretable due to the explicit modeling of information flow through edge connections. We propose the first application of graph neural networks to seasonal forecasting. We design a novel graph connectivity learning module that enables our GNN model to learn large-scale spatial interactions jointly with the actual ENSO forecasting task. Our model, \graphino, outperforms state-of-the-art deep learning-based models for forecasts up to six months ahead. Additionally, we show that our model is more interpretable as it learns sensible connectivity structures that correlate with the ENSO anomaly pattern.

cs.LG↗

arXiv2021

Graph Neural Networks for Improved El Niño Forecasting

Deep learning-based models have recently outperformed state-of-the-art seasonal forecasting models, such as for predicting El Niño-Southern Oscillation (ENSO). However, current deep learning models are based on convolutional neural networks which are difficult to interpret and can fail to model large-scale atmospheric patterns called teleconnections. Hence, we propose the application of spatiotemporal Graph Neural Networks (GNN) to forecast ENSO at long lead times, finer granularity and improved predictive skill than current state-of-the-art methods. The explicit modeling of information flow via edges may also allow for more interpretable forecasts. Preliminary results are promising and outperform state-of-the art systems for projections 1 and 3 months ahead.

cs.LG↗