Deshpande Shachi, Marx Charles, Kuleshov Volodymyr
Cornell Tech and Cornell University.
Stanford University.
Proc Mach Learn Res. 2024 May;238:1450-1458.
Accurate uncertainty estimates are important in sequential model-based decision-making tasks such as Bayesian optimization. However, these estimates can be imperfect if the data violates assumptions made by the model (e.g., Gaussianity). This paper studies which uncertainties are needed in model-based decision-making and in Bayesian optimization, and argues that uncertainties can benefit from calibration-i.e., an 80% predictive interval should contain the true outcome 80% of the time. Maintaining calibration, however, can be challenging when the data is non-stationary and depends on our actions. We propose using simple algorithms based on online learning to provably maintain calibration on non-i.i.d. data, and we show how to integrate these algorithms in Bayesian optimization with minimal overhead. Empirically, we find that calibrated Bayesian optimization converges to better optima in fewer steps, and we demonstrate improved performance on standard benchmark functions and hyperparameter optimization tasks.
在诸如贝叶斯优化等基于序列模型的决策任务中，准确的不确定性估计非常重要。然而，如果数据违反了模型所做的假设（例如高斯性），这些估计可能并不完美。本文研究了基于模型的决策和贝叶斯优化中需要哪些不确定性，并认为不确定性可以从校准中受益——即80%的预测区间应在80%的时间内包含真实结果。然而，当数据是非平稳的且取决于我们的行动时，保持校准可能具有挑战性。我们提出使用基于在线学习的简单算法来可证明地在非独立同分布数据上保持校准，并展示如何以最小的开销将这些算法集成到贝叶斯优化中。从经验上看，我们发现校准后的贝叶斯优化能在更少的步骤中收敛到更好的最优解，并且我们在标准基准函数和超参数优化任务上展示了改进的性能。
