题目：Random Forests and Deep Neural Networks for Euclidean and non-Euclidean regression
报告人：於州（华东师范大学 经济管理学院）
时间：4月21日（星期五），11:00-12:00
地点：明德楼，报告厅B201-1
摘要：Neural networks and random forests are popular and promising tools formachine learning. We explore the proper integration of these two approaches fornonparametric regression to improve the performance of a single approach.Itnaturally synthesizes the local relation adaptivity of random forests and the strongglobal approximation ability of neural networks. By utilizing advanced U-processtheory and an appropriate network structure, we obtain the minimax convergence ratefor the estimator. Moreover, we propose the novel random forest weighted localFrechet regression paradigm for regression with non-Euclidean responses. Weestablish the consistency, rate of convergence, and asymptotic normality for thenon-Euclidean random forests based estimator.
