Q

Definition: A model-free reinforcement learning algorithm that learns the value of an action in a particular state of a Markov decision process (MDP). It uses action-reward learning to provide agents with values that they can use to maximise the total reward.

Definition: In reinforcement learning, the Q-value or action-value represents the expected future rewards that can be obtained by taking a given action from a given state, following a certain policy.

Definition: Another term for the action-value function used in reinforcement learning that maps state-action pairs to rewards. It helps in determining the best action to take under a given state.

Definition: A statistical technique used to separate measurements of two or more classes of objects or events by a quadratic surface. It is similar to linear discriminant analysis but allows for non-linear separation of data.

Definition: Data that approximates or characterises but does not measure the attributes, characteristics, properties, etc., of a thing or phenomenon. Qualitative data describes qualities or characteristics that are described in terms of categories or labels.

Definition: In the context of AI, quality assurance involves the processes and methodologies used to ensure that the developed models and algorithms meet the required standards and function as expected.

Definition: A type of regression analysis used in statistics and machine learning that estimates the quantiles (or percentiles) of the response variable. This method is useful for cases when the conditions of ordinary least squares regression are not applicable or give biased results.

Definition: In digital processing and machine learning, quantisation refers to the process of mapping input values from a large set (often continuous values) to output values in a (countable) smaller set. In neural networks, quantisation reduces the precision of the network’s parameters to reduce memory and improve model inference speed.

Definition: A type of computing that uses quantum-mechanical phenomena, such as superposition and entanglement, to perform operations on data. Quantum computing is anticipated to significantly impact AI by potentially speeding up certain computations, such as data processing and optimisation problems.

Definition: In the context of databases and information retrieval, a query is a request for information from a database. In AI, queries are often used to retrieve data needed for analysis and decision-making.

Definition: A linear data structure or a collection in which the entities are kept in order and the principal operations on the collection are the addition of entities to the rear terminal position (enqueue) and removal of entities from the front terminal position (dequeue). In AI systems, queues might manage tasks or data processing sequences.

Definition: A web-based game developed by Google that uses neural network artificial intelligence to recognise doodles and match them to objects. It serves as an experiment in machine learning and data collection for training AI models.

Definition: In AI, this term describes a state or condition of inactivity or dormancy. In neural networks, for example, a quiescent state might refer to neurons that are currently inactive because their input is below the threshold level required to activate them.

Definition: An algorithm invented by Ross Quinlan used to generate a decision tree based on a set dataset. ID3 stands for Iterative Dichotomiser 3 and is one of the simplest forms of decision tree algorithms used in machine learning.

Definition: A minimum number of members of an ensemble or group that must be present to make the proceedings of that group valid. In distributed computing and AI, achieving a quorum involves having enough nodes in a network agree on a value or decision before it can be accepted as the consensus.