Abstract:In this work we study the estimation of the density of a totally positive random vector. Total positivity of the distribution of a random vector implies a strong form of positive dependence between its coordinates and, in particular, it implies positive association. Since estimating a totally positive density is a non-parametric problem, we take on a (modified) kernel density estimation approach. Our main result is that the sum of scaled standard Gaussian bumps centered at a min-max closed set provably yields a totally positive distribution. Hence, our strategy for producing a totally positive estimator is to form the min-max closure of the set of samples, and output a sum of Gaussian bumps centered at the points in this set. We can frame this sum as a convolution between the uniform distribution on a min-max closed set and a scaled standard Gaussian. We further conjecture that convolving any totally positive density with a standard Gaussian remains totally positive.

On the completely positive kernels for nonuniform meshes

A class of monotonicity-preserving variable-step discretizations for Volterra integral equations

A class of monotonicity-preserving variable-step discretizations for Volterra integral equations and time fractional ordinary differential equations

Nonnegativity preserving convolution kernels. Application to Stochastic Volterra Equations in closed convex domains and their approximation

Strictly positive definite kernels on two-point compact homogeneous spaces

SYMMETRY AND NONEXISTENCE OF POSITIVE SOLUTIONS TO FRACTIONAL P-LAPLACIAN EQUATIONS

UNIFORM POLYNOMIAL STABILITY OF SECOND ORDER INTEGRO-DIFFERENTIAL EQUATIONS IN HILBERT SPACES WITH POSITIVE DEFINITE KERNELS

Positive definite kernels and boundary spaces

Regularity and Convergence Results for Nonlocal Peridynamic Equations with Truncated Tensor Kernels

Positivity and Monotonicity Results for Discrete Fractional Operators Involving the Exponential Kernel.

Noncommutative reproducing kernel Hilbert spaces

On The Integral Equation With The Axis-Symmetric Kernel

Some Grönwall inequalities for a class of discretizations of time fractional equations on nonuniform meshes

Hyperdeterminantal Total Positivity

WELL-POSEDNESS AND CONVERGENCE ANALYSIS OF A NONLOCAL MODEL WITH SINGULAR MATRIX KERNEL

Convolutions of Totally Positive Distributions with applications to Kernel Density Estimation

Asymptotically Compatible Fractional Grönwall Inequality and its Applications

Propagation of regularity and positive definiteness: a constructive approach

Positive definite singular kernels on two-point homogeneous spaces

FEM on nonuniform meshes for nonlocal Laplacian: Semi-analytic Implementation in One Dimension

Generalized Convolution Roots of Positive Definite Kernels on Complex Spheres