Nearly-minimax variance estimation under rough random design

{"We":[0],"identify":[1],"the":[2,30,37,55,69],"minimax":[3,56],"exponent":[4],"for":[5],"constant":[6],"conditional":[7,31],"variance":[8],"estimation":[9],"under":[10],"rough":[11],"random":[12],"design.":[13],"The":[14],"unknown":[15],"design":[16],"density":[17],"is":[18,74,87],"bounded":[19,42],"above":[20],"and":[21,29,39,63,84],"away":[22],"from":[23],"zero,":[24],"with":[25,50],"no":[26],"smoothness":[27],"assumption,":[28],"error":[32],"laws":[33],"may":[34],"depend":[35],"on":[36],"covariates":[38],"have":[40],"uniformly":[41,76],"fourth":[43],"moments.":[44],"For":[45,82],"an":[46],"$s$-H\\\\\\"older":[47],"regression":[48],"function":[49],"$s>1$":[51],"in":[52],"dimension":[53],"$d>4s$,":[54],"root-mean-square":[57],"risk":[58],"lies":[59],"between":[60],"$cn^{-2(s+1)/(d+4)}e^{-C\\\\sqrt{\\\\log":[61],"n}}$":[62],"$Cn^{-2(s+1)/(d+4)}$.":[64],"These":[65],"bounds":[66],"show":[67],"that":[68],"rate":[70],"proposed":[71],"by":[72],"Robins":[73],"not":[75],"attainable":[77],"over":[78],"this":[79],"model":[80],"class.":[81],"$01$":[83],"$d\\\\le4s$,":[85],"it":[86],"$n^{-1/2}$.":[88]}