1,721,125 research outputs found
A modal approach for clustering matrices
No full textIn this work we propose a modal approach to density-based clustering for
matrix-valued data. We introduce appropriate kernels for this type of data and define
a kernel estimator of matrix-variate density functions. Additionally, we propose an
extension of the mean-shift procedure for the identification of the modes of the estimated density. Given the intrinsic high dimensionality of matrix-variate data and the
resulting computational complexity of the algorithm, we discuss a possible solution
to handle the problem. We finally present the performance of the proposed method
through an application to real world data, also with respect to a plausible competitor
Nonparametric tests for semiparametric regression models
Full text linkSemiparametric regression models have received considerable attention over the last decades, because of their flexibility and their good finite sample performances. Here we propose an innovative nonparametric test for the linear part of the models, based on random sign-flipping of an appropriate transformation of the residuals, that exploits a spectral decomposition of the residualizing matrix associated with the nonparametric part of the model. The test can be applied to a vast class of extensively used semiparametric regression models with roughness penalties, with nonparametric components defined over one-dimensional, as well as over multi-dimensional domains, including, for instance, models based on univariate or multivariate splines. We prove the good asymptotic properties of the proposed test. Moreover, by means of extensive simulation studies, we show the superiority of the proposed test with respect to current parametric alternatives, demonstrating its excellent control of the Type I error, accompanied by a good power, even in challenging data scenarios, where instead current parametric alternatives fail
Identification of high-energy γ−ray sources via nonparametric clustering
No full textHigh-energy gray sources exhibit as energy flares in the sky map,
produced by variously concentrated photon emissions. The identification of these
sources is a fundamental task to better understand the mechanisms that both create and accelerate particles emitted by celestial objects. We discuss the application
of nonparametric clustering for gray source detection and provide an algorithm
specific for this task. The procedure accounts for the intrinsic uncertainty associated to the available data, arising as an effect of the instrument-pitch and multiple
scattering
A high-resolution aeromagnetic field test in Friuli: towards developing remote location of buried ferro-metallic bodies
No full textHigh Resolution AeroMagnetic surveys (HRAM) are a novel tool experimented in several countries for volcano
and earthquake hazard re-assessment, ground water exploration and mitigation, hazardous waste site characterization
and accurate location of buried ferrous objects (drums, UXO, pipelines). The improvements achieved by HRAM
stem from lower terrain clearance coupled with accurately positioned, real-time differential navigation on closely
spaced flight grids. In field cultural noise filtering, advanced data processing, imaging and improved interpretation
techniques enhance data information content. Development of HRAM approaches might also contribute to mitigate
environmental hazards present throughout the Italian territory. Hence an HRAM field test was performed in July
2000 in Friuli, North-Eastern Italy to assess the capabilities and limitations of HRAM over a buried pipeline and
a domestic waste site. A Cesium magnetometer in towed bird configuration was used on two separate grids.
Profile line spacing was 50-100 m and bird nominal ground clearance was set to 50 m. Microlevelled total field
magnetic anomaly data forms the basis for subsequent advanced processing products including 3D analytic signal,
maximum horizontal gradient of pseudo-gravity and 3D Euler Deconvolution. The magnetic signatures we detected
and enhanced over the environmental test site area in Friuli are also compared with similar but more extensive
HRAM signatures recently observed in other countries
Analysis of Data Over Complex Regions
No full textWe present nonparametric method for data distributed over complex spatial domains. In particular, we consider hypothesis testing procedures in the case of spatial regression models with dif- ferential regularization. We also consider a nonparametric penalized likelihood approach to density estimation over planar domains with complex geometry. The model formulation is based on a regular- ization with differential operators and it is made computationally tractable by means of finite element method. The performances of the proposed methods are presented through extended simulation studies
Bounded Domain Density Estimation
No full textIn this work we present a nonparametric penalized likelihood approach to
density estimation. We consider planar domains with complex geometry, characterized by nonlinear boundaries and interior holes. The model formulation is based on
a regularization with differential operators and it is made computationally tractable
by means of finite element method
- …
