Stochastic Optimization Schemes for Performative Prediction with Nonconvex Loss
Abstract: This paper studies a risk minimization problem with decision dependent data distribution. The problem pertains to the performative prediction setting in which a trained model can affect the outcome estimated by the model. Such dependency creates a feedback loop that influences the stability of optimization algorithms such as stochastic gradient descent (SGD). We present the first study on performative prediction with smooth but possibly non-convex loss. We analyze a greedy deployment scheme with SGD (SGD-GD). Note that in the literature, SGD-GD is often studied with strongly convex loss. We first propose the definition of stationary performative stable (SPS) solutions through relaxing the popular performative stable condition. We then prove that SGD-GD converges to a biased SPS solution in expectation. We consider two conditions of sensitivity on the distribution shifts: (i) the sensitivity is characterized by Wasserstein-1 distance and the loss is Lipschitz w.r.t.~data samples, or (ii) the sensitivity is characterized by total variation (TV) divergence and the loss is bounded. In both conditions, the bias levels are proportional to the stochastic gradient's variance and sensitivity level. Our analysis is extended to a lazy deployment scheme where models are deployed once per several SGD updates, and we show that it converges to an SPS solution with reduced bias. Numerical experiments corroborate our theories.
- Manipulation-proof machine learning. arXiv preprint arXiv:2004.03865, 2020.
- Performative prediction in a stateful world. In International conference on artificial intelligence and statistics, pages 6045–6061. PMLR, 2022.
- Optimum statistical estimation with strategic data sources. In Conference on Learning Theory, pages 280–296. PMLR, 2015.
- Stochastic optimization under distributional drift. Journal of machine learning research, 24(147):1–56, 2023.
- Adversarial classification. In Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining, pages 99–108, 2004.
- Stochastic approximation beyond gradient for signal processing and machine learning. IEEE Transactions on Signal Processing, 2023.
- Stochastic optimization with decision-dependent distributions. Mathematics of Operations Research, 48(2):954–998, 2023.
- Nonconvex online support vector machines. IEEE Transactions on Pattern Analysis and Machine Intelligence, 33(2):368–381, 2010.
- Stochastic first-and zeroth-order methods for nonconvex stochastic programming. SIAM journal on optimization, 23(4):2341–2368, 2013.
- On choosing and bounding probability metrics. International statistical review, 70(3):419–435, 2002.
- Strategic classification. In Proceedings of the 2016 ACM conference on innovations in theoretical computer science, pages 111–122, 2016.
- Spambase. UCI Machine Learning Repository, 1999. DOI: https://doi.org/10.24432/C53G6X.
- How to learn when data reacts to your model: performative gradient descent. In International Conference on Machine Learning, pages 4641–4650. PMLR, 2021.
- Regret minimization with performative feedback. In International Conference on Machine Learning, pages 9760–9785. PMLR, 2022.
- State dependent performative prediction with stochastic approximation. In International Conference on Artificial Intelligence and Statistics, pages 3164–3186. PMLR, 2022.
- Multi-agent performative prediction with greedy deployment and consensus seeking agents. Advances in Neural Information Processing Systems, 35:38449–38460, 2022.
- Two-timescale derivative free optimization for performative prediction with markovian data. ArXiv preprint arXiv:2310.05792, 2023.
- Stochastic optimization for performative prediction. Advances in Neural Information Processing Systems, 33:4929–4939, 2020.
- Outside the echo chamber: Optimizing the performative risk. In International Conference on Machine Learning, pages 7710–7720. PMLR, 2021.
- Performative prediction with neural networks. In International Conference on Artificial Intelligence and Statistics, pages 11079–11093. PMLR, 2023.
- Multiplayer performative prediction: Learning in decision-dependent games. Journal of Machine Learning Research, 24(202):1–56, 2023.
- Performative prediction. In International Conference on Machine Learning, pages 7599–7609. PMLR, 2020.
- Multi-agent performative prediction: From global stability and optimality to chaos. In Proceedings of the 24th ACM Conference on Economics and Computation, pages 1047–1074, 2023.
- Dataset shift in machine learning. Mit Press, 2022.
- Decision-dependent risk minimization in geometrically decaying dynamic environments. In Proceedings of the AAAI Conference on Artificial Intelligence, pages 8081–8088, 2022.
- Constrained stochastic nonconvex optimization with state-dependent markov data. In Advances in Neural Information Processing Systems, volume 35, pages 23256–23270, 2022.
- Network effects in performative prediction games. In International Conference on Machine Learning, pages 36514–36540. PMLR, 2023.
- Stochastic saddle point problems with decision-dependent distributions. SIAM Journal on Optimization, 33(3):1943–1967, 2023.
- Yulai Zhao. Optimizing the performative risk under weak convexity assumptions. arXiv preprint arXiv:2209.00771, 2022.
- Online performative gradient descent for learning nash equilibria in decision-dependent games. Advances in Neural Information Processing Systems, 36, 2023.
- Who leads and who follows in strategic classification? In Advances in Neural Information Processing Systems, volume 34, pages 15257–15269, 2021.
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.