Performative Time-Series Forecasting
Abstract: Time-series forecasting is a critical challenge in various domains and has witnessed substantial progress in recent years. Many real-life scenarios, such as public health, economics, and social applications, involve feedback loops where predictions can influence the predicted outcome, subsequently altering the target variable's distribution. This phenomenon, known as performativity, introduces the potential for 'self-negating' or 'self-fulfilling' predictions. Despite extensive studies in classification problems across domains, performativity remains largely unexplored in the context of time-series forecasting from a machine-learning perspective. In this paper, we formalize performative time-series forecasting (PeTS), addressing the challenge of accurate predictions when performativity-induced distribution shifts are possible. We propose a novel approach, Feature Performative-Shifting (FPS), which leverages the concept of delayed response to anticipate distribution shifts and subsequently predicts targets accordingly. We provide theoretical insights suggesting that FPS can potentially lead to reduced generalization error. We conduct comprehensive experiments using multiple time-series models on COVID-19 and traffic forecasting tasks. The results demonstrate that FPS consistently outperforms conventional time-series forecasting methods, highlighting its efficacy in handling performativity-induced challenges.
- Bartlett, P. L. 1992. Learning with a slowly changing distribution. In Proceedings of the fifth annual workshop on Computational learning theory, 243–252.
- Performative Prediction in Time Series: A Case Study. In NeurIPS 2022 Workshop on Learning from Time Series for Health.
- Performative prediction in a stateful world. In International Conference on Artificial Intelligence and Statistics, 6045–6061. PMLR.
- Alignment of noisy signals. IEEE Transactions on Instrumentation and Measurement, 50(1): 141–149.
- Strategic classification in the dark. In International Conference on Machine Learning, 3672–3681. PMLR.
- Efficiently Modeling Long Sequences with Structured State Spaces. In The International Conference on Learning Representations (ICLR).
- Strategic classification. In Proceedings of the 2016 ACM conference on innovations in theoretical computer science, 111–122.
- Healy, K. 2015. The performativity of networks. European Journal of Sociology/Archives Européennes de Sociologie, 56(2): 175–205.
- Forecasting: principles and practice. OTexts.
- CAMul: Calibrated and Accurate Multi-view Time-Series Forecasting. In Proceedings of the ACM Web Conference 2022, 3174–3185.
- Learning time-varying concepts. Advances in neural information processing systems, 3.
- Kulynych, B. 2022. Causal prediction can induce performative stability. In ICML 2022: Workshop on Spurious Correlations, Invariance and Stability.
- An iterative algorithm for joint signal and interference alignment. In 2010 IEEE International Symposium on Information Theory, 2293–2297. IEEE.
- Generalization bounds for non-stationary mixing processes. Machine Learning, 106(1): 93–117.
- Modeling long-and short-term temporal patterns with deep neural networks. In The 41st international ACM SIGIR conference on research & development in information retrieval, 95–104.
- Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting. In International Conference on Learning Representations (ICLR ’18).
- Time-series forecasting with deep learning: a survey. Philosophical Transactions of the Royal Society A, 379(2194): 20200209.
- Learning under concept drift: A review. IEEE transactions on knowledge and data engineering, 31(12): 2346–2363.
- Do economists make markets?: on the performativity of economics. Princeton University Press.
- Performative Reinforcement Learning. arXiv preprint arXiv:2207.00046.
- Stochastic optimization for performative prediction. Advances in Neural Information Processing Systems, 33: 4929–4939.
- Strategic classification is causal modeling in disguise. In International Conference on Machine Learning, 6917–6926. PMLR.
- Rademacher complexity bounds for non-iid processes. Advances in Neural Information Processing Systems, 21.
- Nadaraya, E. A. 1964. On estimating regression. Theory of Probability & Its Applications, 9(1): 141–142.
- A shape-based similarity measure for time series data with ensemble learning. Pattern Analysis and Applications, 16: 535–548.
- A Time Series is Worth 64 Words: Long-term Forecasting with Transformers. arXiv preprint arXiv:2211.14730.
- N-BEATS: Neural basis expansion analysis for interpretable time series forecasting. In International Conference on Learning Representations.
- Performative prediction. In International Conference on Machine Learning, 7599–7609. PMLR.
- EINNs: Epidemiologically-Informed Neural Networks. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37.
- Data-centric epidemic forecasting: A survey. arXiv preprint arXiv:2207.09370.
- DeepAR: Probabilistic forecasting with autoregressive recurrent networks. International Journal of Forecasting, 36(3): 1181–1191.
- A tutorial on support vector regression. Statistics and computing, 14(3): 199–222.
- Deep learning for time series forecasting: a survey. Big Data, 9(1): 3–21.
- Attention is all you need. Advances in neural information processing systems, 30.
- Gaussian processes for regression. Advances in neural information processing systems, 8.
- Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting. Advances in Neural Information Processing Systems, 34: 22419–22430.
- Urban computing: concepts, methodologies, and applications. ACM Transactions on Intelligent Systems and Technology (TIST), 5(3): 1–55.
- Informer: Beyond efficient transformer for long sequence time-series forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, 11106–11115.
- Statstream: Statistical monitoring of thousands of data streams in real time. In VLDB’02: Proceedings of the 28th International Conference on Very Large Databases, 358–369. Elsevier.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.
