應理學院應用數學係邀請,美國西喬治亞大學魏鳳榮博士將於2018年6月4日-6月5日訪問我校並做學術報告。
題目:Sparse structure selection and estimation across and within variables in nonparametric additive model
時間:6月4日周一下午3:00
地點:曲江9-313室
報告摘要:The sparse structure refers to two types of sparsity in this paper, sparsity across variables and sparsity within variables. Sparsity across variables corresponds to irrelevant components in the additive model, which are zero functions over the whole domain; sparsity within variables corresponds to zero sub-regions within the domain of relevant components, which means the underlying important function of the component is zero on some regions of the domain. To select and estimate the sparse structure, I approximate each component by B-spline basis functions and propose a group bridge penalized method which can simultaneously identify the zero functions and zero structure within the nonzero functions. Under appropriate conditions, both convergence and sparse consistency are obtained. Simulation studies indicate that the proposed group bridge penalized B-spline method has superior performance in function and within function sparsity selection and estimation.
報告人簡介:魏鳳榮博士,於2009年在愛荷華大學獲得博士學位,並於2014年獲得西喬治亞大學的終身副教授職位。其研究方向是,金融,網絡分析,模型選擇,高維數據分析,參數、半參數和非參數模型,以及統計遺傳學。共參加主持科研項目六項,合計經費共$60,000。在包括Annals of Statistics, Bernoulli等SCI期刊中,共發表文章十多篇。
歡迎廣大師生踴躍參加!
理 學 院
2018年6月2日