paper-with-me

홈 › Papers

Calibration of Time-Series Forecasting: Detecting and Adapting Context-Driven Distribution Shift

2023-10-23 · Mouxiang Chen, Lefei Shen, Han Fu, Zhuo Li, Jianling Sun, Chenghao Liu

Recent years have witnessed the success of introducing deep learning models to time series forecasting. From a data generation perspective, we illustrate that existing models are susceptible to distribution shifts driven by temporal contexts, whether observed or unobserved. Such context-driven distribution shift (CDS) introduces biases in predictions within specific contexts and poses challenges for conventional training paradigms. In this paper, we introduce a universal calibration methodology for the detection and adaptation of CDS with a trained model. To this end, we propose a novel CDS detector, termed the "residual-based CDS detector" or "Reconditionor", which quantifies the model's vulnerability to CDS by evaluating the mutual information between prediction residuals and their corresponding contexts. A high Reconditionor score indicates a severe susceptibility, thereby necessitating model adaptation. In this circumstance, we put forth a straightforward yet potent adapter framework for model calibration, termed the "sample-level contextualized adapter" or "SOLID". This framework involves the curation of a contextually similar dataset to the provided test sample and the subsequent fine-tuning of the model's prediction layer with a limited number of steps. Our theoretical analysis demonstrates that this adaptation strategy can achieve an optimal bias-variance trade-off. Notably, our proposed Reconditionor and SOLID are model-agnostic and readily adaptable to a wide range of models. Extensive experiments show that SOLID consistently enhances the performance of current forecasting models on real-world datasets, especially on cases with substantial CDS detected by the proposed Reconditionor, thus validating the effectiveness of the calibration approach.

📄 PDF Abstract BibTeX arXiv:2310.14838

Code (2)

half111/calibration_cds 공식 구현 pytorch
sjtu-dmtai/onlinetsf pytorch

Tasks

Time SeriesTime Series Forecasting

Methods 이 논문이 사용한 방법론

Multi-Head Attention 설명 없음
Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Position-Wise Feed-Forward Layer 설명 없음
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…
Adam 설명 없음

Similar Papers 제목 키워드 기반

Adaptive Multi-Scale Forecasting and Gate-Localized Conformal Prediction for Multivariate Nonstationary Time Series

2026-07-25 · Ziling Ma, Junshu Jiang, Ángel López-Oriona, Ying Sun 외 arxiv

We propose ABF-T-GLCP, a model-agnostic framework for forecasting and uncertainty quantification in nonstationary multivariate time series. The central idea is to learn an adaptive predictive state representation for poi…

Rethinking Time Series Forecasting with LLMs via Nearest Neighbor Contrastive Learning

2024-12-06 · Jayanie Bogahawatte, Sachith Seneviratne, Maneesha Perera, Saman Halgamuge

Adapting Large Language Models (LLMs) that are extensively trained on abundant text data, and customizing the input prompt to enable time series forecasting has received considerable attention. While recent work has show…

Contrastive LearningTime SeriesTime Series Forecasting

Future-Guided Learning: A Predictive Approach To Enhance Time-Series Forecasting

2024-10-19 · Skye Gunasekaran, Assel Kembay, Hugo Ladret, Rui-Jie Zhu 외

Accurate time-series forecasting is essential across a multitude of scientific and industrial domains, yet deep learning models often struggle with challenges such as capturing long-term dependencies and adapting to drif…

Seizure predictionTime SeriesTime Series Forecasting

Perceive, Route and Modulate: Dynamic Pattern Recalibration for Time Series Forecasting

2026-05-07 · Siru Zhong, Zhao Meng, Haohuan Fu, Haoyang Li 외 arxiv

Local temporal patterns in real-world time series continuously shift, rendering globally shared transformations suboptimal. Current deep forecasting models, despite their scale and complexity, rely on fixed weight matric…

Time Series Forecasting

An Interpretable Generative Framework for Anomaly Detection in High-Dimensional Financial Time Series

2026-03-09 · Waldyn G Martinez arxiv

Detecting structural instability and anomalies in high-dimensional financial time series is challenging due to complex temporal dependence and evolving cross-sectional structure. We propose ReGEN-TAD, an interpretable ge…

Anomaly Detection