پیوندهای مفید
آمار
| تعداد نشریات | 32 |
| تعداد شماره‌ها | 604 |
| تعداد مقالات | 5,906 |
| تعداد مشاهده مقاله | 8,922,561 |
| تعداد دریافت فایل اصل مقاله | 6,535,777 |
Robust Estimation of Moment Functions From Discretely Sampled Functional Data | ||
| Caspian Journal of Mathematical Sciences | ||
| مقالات آماده انتشار، پذیرفته شده، انتشار آنلاین از تاریخ 21 آبان 1405 | ||
| نوع مقاله: Research Articles | ||
| شناسه دیجیتال (DOI): 10.22080/cjms.2026.32177.1876 | ||
| نویسندگان | ||
| behdad mostafaiy* ؛ Nahideh Mousavian Koohsareh؛ Masoud Ganji | ||
| University of Mohaghegh Ardabili | ||
| تاریخ دریافت: 31 خرداد 1405، تاریخ بازنگری: 20 مرداد 1405، تاریخ پذیرش: 27 مرداد 1405 | ||
| چکیده | ||
| We consider the problem of estimating higher order moment functions of a stochastic process from discretely observed and noisy functional data. These nonlinear functionals are highly sensitive to measurement error and outliers, requiring robust methods for reliable inference. We propose a unified approach based on bias corrected pseudo observations and robust smoothing spline estimation in a reproducing kernel Hilbert space. The method accommodates both common and independent sampling designs and remains stable under contaminated or heavy tailed noise through M-type estimation. Simulation studies under Gaussian and contaminated Gaussian settings demonstrate that the proposed estimator is competitive with classical methods under clean data and substantially more robust under contamination, with performance depending on the sampling design. The proposed method is also illustrated using the Tecator near-infrared spectrometric data. | ||
| کلیدواژه‌ها | ||
| Functional data analysis؛ Robust estimation؛ Moment function؛ M-type estimation | ||
|
آمار تعداد مشاهده مقاله: 0 |
||