Probability Distributions & Second-Moment Structures

← Back to Domain-Specific Families

Abstractions about probability distributions and dependence structure, covering distribution families such as location-scale, noncentral and reciprocal distributions, second-moment and covariance tools like covariance operators, whitening transformations and uncorrelatedness, and estimation methods such as U-statistics, functional principal component analysis, and control variates.

32 abstractions in this family — domain-specific abstractions that sit near one another in structural-signature space (k-means over structural-signature embeddings). Each is shown with its short description.

• Complex random vector — A random element of a finite-dimensional complex vector space, equivalently a jointly distributed collection of complex-valued random variables.
• Control variates — A Monte Carlo variance-reduction method that adjusts an estimator using correlated quantities whose expectations are known.
• Covariance operator — The linear operator encoding second-order variation of a random element by mapping a direction to its expected covariance-weighted displacement.
• Empirical process — A stochastic process indexing the centered and scaled difference between an empirical measure and its population expectation over a class of functions or sets.
• Energy distance — A metric between probability distributions built from expected pairwise Euclidean distances within and across independent samples.
• Epidemic models on lattices — Spatial epidemic models that place epidemiological states on lattice sites and restrict transmission or state change through local neighborhood rules.
• Exchangeable random variables — A finite or infinite sequence whose joint probability law is invariant under every finite permutation of its indices.
• Fisher information — The expected squared score, or negative expected log-likelihood curvature under regularity conditions, measuring local sensitivity of a probability model to its parameter.
• Folded-t and half-t distributions — Nonnegative distributions obtained by taking the absolute value of a Student-t variate, with the half-t arising from a centered symmetric t distribution restricted or folded at zero.
• Functional correlation — A family of dependence measures for paired random functions that reduces infinite-dimensional covariance structure to interpretable associations between curves or functional components.
• Functional principal component analysis — A dimension-reduction method representing random curves or functions in the eigenbasis of their covariance operator.
• Gaussian process emulator — A probabilistic surrogate that uses a Gaussian process fitted to selected simulator runs to predict an expensive model's output and quantify interpolation uncertainty.
• Invariant estimator — An estimator whose output transforms compatibly with a group action applied to both data and parameter space.
• K-statistic — A symmetric unbiased estimator of a population cumulant constructed from sample power sums.
• Kernel smoother — A nonparametric estimator that predicts a function by distance-weighted averaging of nearby observations.
• Law of total covariance — The identity decomposing covariance into expected conditional covariance plus covariance of conditional expectations.
• Linear belief function — A Dempster–Shafer belief-function representation for continuous variables in which evidence is encoded by linear equations with normal residual uncertainty.
• Location parameter — A distribution parameter whose change translates the probability law along its sample space without changing its shape.
• Location–scale family — A family of probability distributions closed under positive affine transformations of a fixed standardized random variable.
• Maximal information coefficient — A normalized statistic searching over bounded grid partitions to measure potentially nonlinear association between two variables.
• Modified half-normal distribution — A positive-support probability family extending the half-normal shape with power and exponential-tilt parameters.
• Noncentral distribution — A distribution family for a statistic under a shifted alternative, indexed by a noncentrality parameter in addition to the central family parameters.
• Nonparametric skew — A bounded skewness statistic comparing a distribution’s mean and median relative to its mean absolute deviation.
• Orthogonality principle — The condition that a minimum-mean-square estimation error is orthogonal to every admissible variation or estimator-measurable function.
• Reciprocal distribution — A bounded positive distribution whose density is proportional to one over the variable, equivalently uniform after logarithmic transformation.
• Regularized canonical correlation analysis — A canonical-correlation method that stabilizes singular or ill-conditioned covariance estimates by adding penalties, commonly ridge terms, before solving for paired linear variates.
• Simon model — A stochastic growth model in which new categories enter at a fixed probability while existing categories receive new occurrences in proportion to their current counts, generating a power-law size distribution.
• Statistical manifold — A differentiable family of probability distributions equipped with information-geometric structures, most canonically the Fisher information metric and compatible affine connections.
• Total variation — A supremum-based measure of the total accumulated magnitude of change in a function, path, signed measure, or related object.
• U-statistic — A statistic formed by averaging a symmetric kernel over all fixed-size subsets of a sample, yielding an unbiased estimator of its corresponding population functional.
• Uncorrelatedness — A second-moment relation in which two random variables or vectors have zero covariance, excluding linear association without generally implying independence.
• Whitening transformation — A linear transformation that maps a centered random vector with nonsingular covariance to variables having identity covariance.