動径基底関数ニューラルネットワークを適用した再生関数のノンパラメトリック推定
The Nonparametric Estimation of the Renewal Function Applying the Radial Basis Function Neural Network

Shuji Nagai, Tadashi Dohi, Shunji Osaki
2000 Nihon Oyo Suri Gakkai ronbunshi  
. This paper I )roposes a new estimation method of the renewal function from i . i、 d. observa − t, iolls of the inter − arrival time, using a radial basis fullction ( RBF ) type ef neural network , The bas三 c idea is to solve numericaUy the discri 匕 ized renewal equation based on the corresponding empirical dist , ribution fllnction to the observa . tions . The RBF neural network is applied to approximate tlle lmderlying empiric 亂 l distributiol1 . Thr()ughout numerical experiments , we show
more » ... riments , we show that the proposed Inethod can estimate the renewal funct 玉 on wlth higller precision than existing statistical mc 七 hods for apattern data .
doi:10.11540/jsiamt.10.3_227 fatcat:qkfqjnqagrbodlcjfwqknjbpyy