Density deconvolution based on wavelets with bounded supports

详细信息    查看全文

• 作者：Pensky ; Marianna
• 关键词：Deconvolution ; Wavelet ; Thresholding
• 刊名：Statistics and Probability Letters
• 出版年：2002
• 出版时间：February 1, 2002
• 年：2002
• 卷：56
• 期：3
• 页码：261-269
• 全文大小：133 K

文摘

Under the assumption that both convolution densities, g and q, have finite degrees of smoothness, we construct a nonlinear wavelet estimator of the unknown density g based on wavelets with bounded supports. We show that this estimator provides local adaptivity to the unknown smoothness of g and, hence, performs better than the estimator based on Meyer-type wavelets if g has irregular behavior.