cialNeuralNetworks [C] / / NonparametricFunctionalEstimationandRelatedTopics. Netherlands: kluw-er, 199l :561-576. CUCKERF, AIES. Bestchoicesforregularizationparam-etersinlearningtheory: onthebias-varianceproblem [J]. FoundationofComputationalMathematics,Moncler rotterdam, 2002,2:413428. (Continued from p. 356) Question 2 Can we find a better way to solve the problems in Note 2,Moncler 2012, so reducing the literature [6] in the coefficient of 12 at the same time will not increase factor of 8. [1] [2] [3] [4] [5] [References] GALSG. SurlestheoremsofapproximationdeWeierstrass [J]. Mathomatica, 1981,Moncler jas,23 (46) :25-30. GALSG, SZABAD () SJ. Onmonotoneanddoublymonotonepolynomialapproximation [J]. ActaMathematicaHungarica, l992 ,59:395-399. XIETF, ZHOUSP. Aremarkonapproximationbymonotonesequencesofpolynomials [J]. ActaMathematicaHungarica,Moncler 2012, l995, 67:119121. Xieting Fan, Zhou Songping. Real function approximation theory [M]. Hangzhou: Zhejiang University Press, 1998:2830. WUXQ. Adevelopmentonapproximationbymonotonesequencesofpolynomials [J]. ApproximationTheoryandItsApplications. 1998,14:98-1O1. [6] [7] [8] [9] [10] [11] Xie Tingfan. Monotonous sequence of polynomial approximation [J]. China Institute of Metrology, 2003,outlet Moncler,14 (2): 134 136. CYBENKOG. Approximationbysuperpositionsofsigmoi-dalfunction [J]. MathematifControl, Signals, andSys-tems ,Moncler 2012,1989,Moncler milano,2:303-314. CHENDB. Degreeofapproximationbysuperpositionsofasigmoidalfunction [J]. ApproximationTheoryandItsApplications, 1993,Moncler piumini,9 (3) :17-28. CHENTP. CHENH, IIURW. ApproximationCapabil-ityinC (R) bymultilayerIeedforwardnetworksandrelatedproblems [J]. IEEETransactionsonNeuralNetworks ,1995,6:25-30. CAOFI,outlet moncler, XIETF, XUZB. TheEstimateforapproxl-mationerrorofneuralnetworks: aconstructiveapproach [J]. Neurocomputing ,2008,71:626-630. HoRNIKK,outlet Moncler, STINCHC () MBEM, WHITEH. Muhilayerfeedforwardnetworksareuniversalapproximators [knife. NeuralNetworks ,1989,2:359-366.
|
