Doubly robust off-policy evaluation - Doubly robust off-policy evaluation estimates the expected reward of a target decision policy from data collected by another policy, combining a reward or...
Downhill simplex - Downhill simplex is a derivative-free optimization method commonly referring to the Nelder - Mead algorithm.
Downward refinement property - The downward refinement property is a property of a planning abstraction hierarchy that allows abstract solutions to be refined into concrete solutions...
Doxing - Doxing, also spelled doxxing, is the deliberate online exposure of someone's identity or personal information without their consent, typically to enable...
DPLL algorithm - The DPLL algorithm is a complete search procedure for deciding whether a propositional-logic formula in conjunctive normal form can be satisfied.
Drift (tracking) - Tracking drift is the progressive departure of a visual tracker's estimated target location or representation from the intended target.
Dropout - Dropout is a neural-network regularization technique that randomly suppresses selected activations during training.
Dropping conditions - Dropping conditions is a generalization operation in rule or concept learning that removes one or more requirements from a conjunctive description.
Dual extended Kalman filter - A dual extended Kalman filter (dual EKF or DEKF) uses two coupled estimation processes to track both a system's changing hidden state and uncertain model...
Dual graph - In constraint satisfaction, a dual graph represents each original constraint as a node and connects two nodes when their constraints share one or more...
Dual norm - The dual norm associated with a norm measures the largest inner product a vector can achieve against any vector in that norm's unit ball.
Dual problem - A dual problem is an optimization problem constructed from another problem, called the primal, to obtain bounds on its best objective value.
Dual total correlation - Dual total correlation (DTC) is an information-theoretic measure of dependence among several random variables.
DualDICE - DualDICE is a reinforcement-learning method for estimating corrections between a target policy's discounted state-action distribution and the distribution...
Duality gap - The duality gap is the difference between the optimal values of a primal optimization problem and its associated dual.
Dueling bandit - A dueling bandit is a sequential learning problem in which the learner chooses pairs of options and receives relative feedback, such as which option was...
Dueling deep Q-network - A dueling deep Q-network is a Q-value network architecture that combines a shared state representation with two output streams: a scalar value stream and an...
Dummy variable - A dummy variable is a numerical indicator, usually coded 1 when a specified condition or category applies and 0 otherwise.
Duncan's new multiple range test - Duncan's new multiple range test is a procedure for comparing ordered group means using a sequence of studentized-range comparisons.
Dunn-Sidák test - The Dunn-Sidák test is a multiple-testing adjustment that assigns a smaller significance threshold to each test in a specified family.
Dunnett's test - Dunnett's test is a multiple-comparison procedure for comparing several treatment means with one designated control mean.
Dunning-Kruger Effect Bias - The Dunning - Kruger effect describes a pattern in which people performing poorly on a particular task overestimate their performance or standing relative...
Duplicate sample - In the survey terminology used by the OECD glossary, a duplicate sample is an additional sample collected alongside a first sample in the same survey under...
Duplicated sample - A duplicated sample, in the OECD glossary's survey terminology, is a sample whose units are investigated twice by different parties of investigators.
Durability (database systems) - Durability in database systems is the property that a committed transaction's effects survive the failures covered by the system's guarantee.
Duration - Duration is the amount of time an action or activity takes between its start and end.
Durbin test - The Durbin test is a rank-based test of treatment effects in a balanced incomplete block design.
Durbin-Watson statistic - The Durbin-Watson statistic measures how strongly successive regression residuals differ relative to their overall squared size.
Durbin - Wu - Hausman test - The Durbin-Wu-Hausman test is a family of econometric tests used to assess whether specified regressors can be treated as exogenous in a model that also has...
Dutch eligibility trace - A Dutch eligibility trace is a recursively updated feature-memory vector used in reinforcement-learning algorithms such as true online TD(lambda).
Dyadic distribution - In source coding, a dyadic distribution is a discrete probability distribution whose positive probabilities are integer powers of one half, such as 1/2, 1/4...
Dykstra's projection algorithm - Dykstra's projection algorithm finds the point nearest to a given starting point within an intersection of closed convex sets.
Dyna-Q - Dyna-Q is a reinforcement-learning approach that integrates direct Q-learning, model learning and planning.
Dyna-Q with exploration bonus (Dyna-Q+) - Dyna-Q with exploration bonus, usually called Dyna-Q+, extends Dyna-Q by encouraging the agent to revisit state-action pairs that have gone untried for a...
Dynamic backtracking - Dynamic backtracking is a constraint-search method that revises a decision responsible for a conflict while preserving other assignments that need not be...
Dynamic Bayesian network - A dynamic Bayesian network is a probabilistic graphical model for variables that evolve across a sequence, commonly over time.
Dynamic Conditional Correlation - Dynamic Conditional Correlation, or DCC, is a class of multivariate time-series models that allows correlations to change over time conditional on past...
Dynamic epistemic logic - Dynamic epistemic logic studies how knowledge and beliefs change when information is communicated or events are observed.
Dynamic factor - A dynamic factor is an unobserved component that evolves over time and helps explain shared movement across several observed time series.
Dynamic Markov compression - Dynamic Markov compression, or DMC, is a lossless compression approach that adaptively models the probabilities of incoming bits using a changing state...
Dynamic mode decomposition - Dynamic mode decomposition, usually abbreviated DMD, is a data-driven method for representing observed evolution through spatial patterns and associated...
Dynamic model - A dynamic model, in the machine-learning usage of the supplied glossary, is a model that is retrained or updated frequently as new data become available.
Dynamic perfect hashing - Dynamic perfect hashing is a method for maintaining a changing dictionary of keys while preserving collision-free placement for the currently stored set...
Dynamic programming - Dynamic programming is a method for solving a problem by combining solutions to smaller subproblems and reusing their results.
Dynamic range - Dynamic range is the ratio between the largest and smallest meaningful signal levels in a specified system or scene.
Dynamic search problems - Dynamic search problems are search problems whose relevant description changes over time, such as changing edge costs, newly discovered obstacles, or a...
Dynamic SQL - Dynamic SQL is SQL whose statement text or structure is determined at runtime rather than being completely fixed in the program beforehand.
Dynamic state - A dynamic state describes the quantities needed to model how a system will evolve under its dynamics.