In this paper, we study MAICE-DP method for ill-posed problems. MAICE-DP method is based on Minimum AIC Estimation proposed by K.Tanabe[7] and Discrepancy Principle proposed by V.A.Morozov[3][6]. AIC proposed by H.Akaike[1][2] is one of the most famous information criterions and MAICE is very effective for data smoothing problems. Moreover, DP is well established theorically. But, in practice, this matching method, MAICE-DP, is not effective in solving ill-posed problems. We show this with the aid of numerical examples.
抄録全体を表示