報(bào)告題目:Modelling function-valued processes with non-separable and/or non-stationary covariance structure
主講人:史建清教授(英國紐卡斯?fàn)柎髮W(xué))
時(shí)間:2020年1月7日(周二)10:00 a.m.
地點(diǎn):北院卓遠(yuǎn)樓305會議室
主辦單位:統(tǒng)計(jì)與數(shù)學(xué)學(xué)院
摘要:Separability of the covariance structure is a common assumption for function-valued processes defined on two- or higher-dimensional domains. This assumption is often made to obtain an interpretable model or due to difficulties in modelling a potentially complex covariance structure, especially in the case of sparse designs. We proposed to use Gaussian processes with flexible parametric covariance kernels which allow interactions between the inputs in the covariance structure. When we use suitable covariance kernels, the leading eigen-surfaces of the covariance operator can explain well the main modes of variation in the functional data, including the interactions between the inputs. The results are demonstrated by simulation studies and by applications to real world data.
主講人簡介:
史建清教授,現(xiàn)為英國紐卡斯?fàn)柎髮W(xué)(Newcastle University)數(shù)學(xué)、統(tǒng)計(jì)與物理學(xué)院教授,從2018年10月起擔(dān)任英國艾倫.圖靈研究院(The Alan Turing Institute, 即英國國家數(shù)據(jù)科學(xué)和人工智能研究院)研究員,目前是英國皇家統(tǒng)計(jì)協(xié)會會士(Fellow of the Royal Statistical Society)、泛華統(tǒng)計(jì)協(xié)會(International Chinese Statistical Association)會員、英國皇家統(tǒng)計(jì)協(xié)會期刊評論文章的客座編委(Guest Associate Editor of Journal of the Royal Statistical Society ),統(tǒng)計(jì)學(xué)雜志Journal of the Royal Statistical Society(Series C,應(yīng)用統(tǒng)計(jì))、Statistical and Probability letters、British Journal of Mathematics & Computer Sciences等的編委(Associate Editor)。
史建清教授于1996年在香港中文大學(xué)獲得統(tǒng)計(jì)學(xué)博士學(xué)位,并先后到英國華威大學(xué)(The University of Warwick)和格拉斯哥大學(xué)(The University of Glasgow)從事博士后研究工作,于2002年入職紐卡斯?fàn)柎髮W(xué),分別于2008年和2015年受邀到劍橋大學(xué)艾薩克·牛頓數(shù)學(xué)科學(xué)研究院(Isaac Newton Institute for Mathematical Sciences, Cambridge University)和美國統(tǒng)計(jì)與應(yīng)用數(shù)學(xué)科學(xué)研究院(Statistical and Applied Mathematical Sciences Institute)進(jìn)行學(xué)術(shù)訪問。
史建清教授的主要研究領(lǐng)域包括函數(shù)型數(shù)據(jù)分析、缺失數(shù)據(jù)及模型不確定性分析、協(xié)方差結(jié)構(gòu)分析和潛變量模型等,已在國際學(xué)術(shù)期刊上發(fā)表多篇高水平論文,包括統(tǒng)計(jì)學(xué)頂級期刊Journal of the American Statistical Association、Journal of the Royal Statistical Society(Series B)和Biometrika,目前已發(fā)表學(xué)術(shù)論文70余篇,在國際權(quán)威出版社Chapman & Hall出版專著1部,即《函數(shù)型數(shù)據(jù)的高斯過程回歸分析》(Gaussian Process Regression Analysis for Functional Data)。此外,史建清教授2011年在期刊Journal of Nonparametric Statistics上發(fā)表的論文Bayesian single-index model using a Gaussian process prior獲得美國統(tǒng)計(jì)協(xié)會頒發(fā)的非參數(shù)統(tǒng)計(jì)主題最佳論文獎,2014年的論文Automatic Assessment of Upper Limb Function During Play of the Action Video Game, Circus Challenge: Validity and Sensitivity to Change獲得美國電氣與電子工程師協(xié)會(IEEE)頒發(fā)的最佳論文獎。史建清教授也曾多次獲得英國自然科學(xué)基金(EPSRC,Engineering And Physical Sciences Research Council)、英國國家醫(yī)學(xué)基金(MRC,Medical Research Council))和英國衛(wèi)生改革挑戰(zhàn)基金(Welcome Trust Health Innovation Challenge Fund)等的資助。