Sample Complexity and Decision-Theoretic Guarantees for Bayesian Model Averaging over Decision Trees with Catalan-Exponential Priors
基于Catalan指数先验的决策树贝叶斯模型平均的样本复杂度和决策理论保证
机构 * School of Computing and Engineering University of Bedfordshire, Luton, UK(计算与工程学院贝德福德郡大学,卢顿,英国)
AI总结 针对具有Dirichlet-Multinomial叶模型和Catalan指数树大小先验的贝叶斯决策树,建立了理性承诺阈值的完整非渐近理论,回答了贝叶斯模型平均权重何时蕴含足够认知信息以证明对平均分布的承诺利用是合理的。
Comments 22 pages, 3 figures, Submitted to the Journal of Machine Learning Research