Sensitivity Analysis in Kernel Principal Component Analysis
نویسندگان
چکیده
In this paper we derive empirical influence functions for features in kernel principal component analysis. Based on the derived influence functions, a sensitivity analysis procedure is proposed for detecting influential objects with respect to each feature, subspace spanned by specified eigenvectors, and configuration of the features of interest. We show the usefulness of the proposed procedure with a numerical example.
منابع مشابه
Object Recognition based on Local Steering Kernel and SVM
The proposed method is to recognize objects based on application of Local Steering Kernels (LSK) as Descriptors to the image patches. In order to represent the local properties of the images, patch is to be extracted where the variations occur in an image. To find the interest point, Wavelet based Salient Point detector is used. Local Steering Kernel is then applied to the resultant pixels, in ...
متن کاملMultivariate Statistical Kernel PCA for Nonlinear Process Fault Diagnosis in Military Barracks
Because of the nonlinear characteristics of monitoring system in military barracks, the traditional KPCA method either have low sensitivity or unable to detect the fault quickly and accurately. In order to make use of higher-order statistics to get more useful information and meet the requirements of real-time fault diagnosis and sensitivity, a new method of fault detection and diagnosis is pro...
متن کاملDetecting influential observations in Kernel PCA
Individual observations can be very influential when performing classical Principal Component Analysis in a Euclidean space. Robust PCA algorithms detect and neutralize such dominating data points. This paper studies robustness issues for PCA in a kernel induced feature space. The sensitivity of Kernel PCA is characterized by calculating the influence function. A robust Kernel PCA method is pro...
متن کاملاثر تاریخ کاشت بر تحمل به تنش سرما در ژنوتیپ های جو پاییزه و بهاره
In order to evaluate cold tolerance of twenty barley genotypes under field conditions, an experiment was carried out in a randomized complete block design at 3 sowing dates of October 5, November 5, and December 5 in Saatlu Agricultural Research Station, West Azarbaijan, Iran, during 2010-11 seasons. Also, another experiment was conducted on the same genotypes based on a completely randomized d...
متن کاملA tree kernel to analyse phylogenetic profiles
MOTIVATION The phylogenetic profile of a protein is a string that encodes the presence or absence of the protein in every fully sequenced genome. Because proteins that participate in a common structural complex or metabolic pathway are likely to evolve in a correlated fashion, the phylogenetic profiles of such proteins are often 'similar' or at least 'related' to each other. The question we add...
متن کامل