Multiplicative utility function - A multiplicative utility function combines utilities for several attributes through a product-based expression that includes interactions between those...
Multitask learning - Multitask learning trains related learning tasks jointly so that information learned for one task can help the others.
Mutual information - Mutual information measures statistical dependence between two random variables by comparing their joint distribution with the distribution they would have...
Neighbor - A neighbor of a node in a simple undirected graph is another node connected to it directly by an edge.
Neighborhood - A neighborhood is a collection of points or nodes treated as local to a reference point under a specified rule.
Network Data Envelopment Analysis - Network Data Envelopment Analysis, or network DEA, evaluates the relative efficiency of organizations while representing their internal stages and the flows...
Network Lasso - Network Lasso is an optimization framework that fits variables or local models at graph nodes while penalizing differences across connected nodes.
Network performance - Network performance describes how well a communication network carries data for a specified task and set of conditions.
Network throughput - Network throughput is the rate at which data passes through a network, link or device under specified conditions.
Networked data - Networked data are observations or local datasets accompanied by a network of relationships between them.
Networked exponential families - Networked exponential families are probabilistic models that assign an exponential-family distribution to each node of a graph and couple the node-level...
Networked federated learning - Networked federated learning is a family of methods that trains personalized local models from decentralized datasets while using a network of relationships...
Node degree - Node degree measures the connections incident to a node in a graph. In a simple undirected graph without self-loops, it equals the number of distinct neighbors.
Noisy-channel coding theorem - The noisy-channel coding theorem states that reliable digital communication is possible over a noisy channel at information rates below its capacity, using...
Non-expansive operator - A non-expansive operator is a mapping that never increases the distance between two inputs, measured using a specified norm or metric.
Nonlinear conjugate gradient method - The nonlinear conjugate gradient method is a family of algorithms for minimizing a smooth function by combining the current negative gradient with...
Norm - A norm is a mathematical function that assigns a nonnegative size to a vector while satisfying three rules: only the zero vector has size zero, scaling a...
Normed space - A normed space is a vector space equipped with a norm: a rule that measures the size of each vector consistently with addition and scalar multiplication.
Nullspace - The nullspace of a matrix A is the set of all vectors x that the matrix maps to the zero vector: Ax=0.
Odds algorithm - The odds algorithm is an optimal stopping rule for selecting the last success in a finite sequence of independent yes-or-no trials with known success...
Online algorithm - An online algorithm processes input as it arrives and makes required decisions without knowing the entire future input sequence.
Online gradient descent - Online gradient descent is a method that updates a decision or model using the gradient of the loss revealed after each round.
Online optimization - Online optimization studies how to make a sequence of decisions when relevant information arrives over time and future costs are not yet known.
Optimistic knowledge gradient - Optimistic knowledge gradient is a sequential allocation policy that chooses the next observation using the most favorable possible immediate improvement in...
Optimization - Optimization is the process of choosing allowed values for decision variables to make a stated objective as good as possible.
Optimization method - An optimization method is an algorithm for finding an exact or approximate solution to a specified optimization problem.
Optimization problem - An optimization problem is a mathematical specification of decisions to be chosen, an objective to minimize or maximize, and the set of permitted choices.
Order of accuracy - Order of accuracy describes how quickly a numerical method's approximation error decreases as its step size or grid spacing becomes smaller.
Ordinary Least Squares - Ordinary Least Squares, usually abbreviated OLS, fits a linear model by choosing coefficients that minimize the sum of squared residuals.
Orthogonality condition - The orthogonality condition characterizes the nearest-point projection onto a linear subspace: the projected point lies in the subspace, and the remaining...
Outage probability - Outage probability is the probability that a communication link fails to meet a specified operating requirement under a stated model of changing conditions.
Overlap - add method - The overlap - add method computes a long signal's convolution by processing shorter input blocks and adding the portions of their outputs that occupy the...
Overlap - save method - The overlap - save method computes a long signal's linear convolution using overlapping input frames and circular convolution.
Parameter space - A parameter space is the set of all permitted values of a model's parameters.
Paranoid algorithm - The paranoid algorithm is a game-tree search approach for games with more than two players.
Partial derivative - A partial derivative measures how a function changes as one input variable changes while its other input coordinates are held fixed.
Pattern search (optimization) - Pattern search is a family of derivative-free optimization methods that look for better solutions by evaluating an objective function at a structured set of...
PDE-constrained optimization - PDE-constrained optimization chooses the best design, control or parameter values while requiring a system to obey specified partial differential equations.
Penalty method - A penalty method tackles constrained optimization by adding a cost for violating constraints to the objective being minimized.
Penalty term - A penalty term is an additional part of an optimization objective that makes specified properties or behaviors less attractive.
Polynomial matrix spectral factorization - Polynomial matrix spectral factorization expresses a suitable matrix-valued spectrum as a polynomial matrix multiplied by its conjugate-transposed...
Positive definite (pd) - A real symmetric matrix is positive definite, abbreviated pd, when its quadratic form is strictly positive for every nonzero vector.
Positive semi-definite (psd) - A real symmetric matrix is positive semi-definite, abbreviated psd, when its quadratic form is nonnegative for every vector.
Posterior (prediction) - In the prediction sense used here, a posterior is the conditional probability distribution of an unknown label or target given observed features.
Powell's dog leg method - Powell's dogleg method is a trust-region optimization technique that selects a step along a two-segment path connecting a steepest-descent-based step to a...
Powell's method - Powell's method is a derivative-free optimization approach that improves a scalar objective through a sequence of one-dimensional searches along selected...
Probabilistic principal component analysis - Probabilistic principal component analysis, or PPCA, is a probabilistic version of PCA that models an observed vector as a linear transformation of a...