A general non-parametric technique is proposed for the analysis of a complex multimodal feature space and to delineate arbitrarily shaped clusters in it. The basic computational module of the technique is an old pattern recognition procedure: the mean shift. For discrete data, we prove the convergence of a recursive mean shift procedure to the nearest stationary point of the underlying density function and, thus, its utility in detecting the modes of the density. The relation of the mean shift procedure to the Nadaraya-Watson estimator from kernel regression and the robust M-estimators; of location is also established. Algorithms for two low-level vision tasks discontinuity-preserving smoothing and image segmentation - are described as applications. In these algorithms, the only user-set parameter is the resolution of the analysis, and either gray-level or color images are accepted as input. Extensive experimental results illustrate their excellent performance.

Mean shift: a robust approach toward feature space analysis

We present a new adaptive kernel density estimator based on linear diffusion processes. The proposed estimator builds on existing ideas for adaptive smoothing by incorporating information from a pilot density estimate. In addition, we propose a new plug-in bandwidth selection method that is free from the arbitrary normal reference rules used by existing methods. We present simulation examples in which the proposed approach outperforms existing methods in terms of accuracy and reliability.

Kernel density estimation via diffusion

/pdf/kernel-density-estimation-via-diffusion-716yd6ge0c.pdf

Summary

This paper considers the problem of selecting optimal bandwidths for variable (sample-point adaptive) kernel density estimation. A data-driven variable bandwidth selector is proposed, based on the idea of approximating the log-bandwidth function by a cubic spline. This cubic spline is optimized with respect to a cross-validation criterion. The proposed method can be interpreted as a selector for either integrated squared error (ISE) or mean integrated squared error (MISE) optimal bandwidths. This leads to reflection upon some of the differences between ISE and MISE as error criteria for variable kernel estimation. Results from simulation studies indicate that the proposed method outperforms a fixed kernel estimator (in terms of ISE) when the target density has a combination of sharp modes and regions of smooth undulation. Moreover, some detailed data analyses suggest that the gains in ISE may understate the improvements in visual appeal obtained using the proposed variable kernel estimator. These numerical studies also show that the proposed estimator outperforms existing variable kernel density estimators implemented using piecewise constant bandwidth functions.

Variable kernel density estimation

Given the increasing interest in agricultural risk, many have sought improved methods to characterize conditional crop-yield densities While most have postulated the Beta as a flexible alternative to the Normal, others have chosen nonparametric methods Unfortunately, yield data tends not to be sufficiently abundant to invalidate many reasonable parametric models This is problematic because conclusions from economic analyses, which require estimated conditional yield densities, tend not to be invariant to the modeling assumption We propose a semiparametric estimator that, because of its theoretical properties and our simulation results, enables one to empirically proceed with a higher degree of confidence Copyright 2003, Oxford University Press

Modeling Conditional Yield Densities

We introduce a data-perturbation method for reducing bias of a wide variety of density estimators, in univariate, multivariate, spatial and spherical data settings. The method involves 'sharpening' the data by making them slightly more clustered than before, and then computing the estimator in the usual way, but from the sharpened data rather than the original data. The transformation depends in a simple, explicit way on the smoothing parameter employed for the density estimator, which may be based on classical kernel methods, orthogonal series, histosplines, singular integrals or other linear or approximately-linear methods. Bias is reduced by an order of magnitude, at the expense of a constant-factor increase in variance.

E Choi

Papers

Data sharpening as a prelude to density estimation