F

Definition: A technology capable of identifying or verifying a person from a digital image or a video frame from a video source by comparing and analysing patterns based on the person’s facial contours.

Definition: A type of error in statistical tests where a test result incorrectly indicates no presence of a condition (the result is negative), when in reality it is present.

Definition: A type of error in which a test result wrongly indicates the presence of a condition (the result is positive), when in reality it is not present.

Definition: In machine learning, a feature is an individual measurable property or characteristic of a phenomenon being observed. Choosing informative, discriminating, and independent features is a crucial element of effective algorithmic models.

Definition: The process of using domain knowledge to select, modify, or create new features from raw data to increase the predictive power of machine learning algorithms.

Definition: A method of dimensionality reduction that involves transforming raw data into a set of features that can be easily processed by a machine learning model.

Definition: The process of identifying and selecting a subset of relevant features for use in model construction. This helps improve the model’s performance and reduces overfitting.

Definition: A machine learning technique where the training algorithm is distributed across multiple decentralised devices or servers without exchanging the data samples. This helps improve privacy and data security.

Definition: A system structure in which the output of a process is used as input back into the same process. This is common in adaptive systems where the system dynamically adjusts its performance based on its output.

Definition: A type of neural network architecture where connections between the nodes do not form a cycle. This is one of the simplest types of artificial neural networks.

Definition: A process in machine learning where a pre-trained model is further trained (tuned) with a smaller dataset for tasks similar but not identical to the ones it was originally trained on.

Definition: Short for financial technology, it refers to new tech that seeks to improve and automate the delivery and use of financial services. AI in fintech is used for tasks like fraud detection, risk management, and customer service.

Definition: Also known as predicate logic, a formalism used in reasoning about relations among objects, and dealing with the notions of properties, relations, and quantification.

Definition: A method used in statistics, pattern recognition, and machine learning to measure the discriminative power of individual features in relation to the outcome variable.

Definition: A particular type of objective function in genetic programming and evolutionary computation that quantifies the optimality of a solution (that is, a chromosome) so that that solution may be ranked against all the other solutions.

Definition: An acronym for “Fast Linear Algebra in Rust,” a tool or library that provides capabilities for natural language processing tasks by using deep learning models.

Definition: A behaviour exhibited by multi-agent systems in which agents, governed by simple rules, operate together to exhibit complex behaviour. This can include simulations of natural phenomena or crowd behavior.

Definition: A diagram that represents a workflow or process. In AI development, flowcharts are used to plan algorithms and illustrate the logic of programs.

Definition: A type of loss function used primarily to address class imbalance during training of a neural network. It applies a modulating term to the cross-entropy loss in order to focus learning on hard misclassified examples.

Definition: An extension of cloud computing which involves the deployment of data, compute, storage, and applications services closer to the end-users or near-user edge devices, like network routers.

Definition: In game theory, a theorem that suggests in infinitely repeated games, any outcome that gives players at least their minimax value can be sustained as a Nash equilibrium.

Definition: A method in artificial intelligence used in rule-based system solvers, where inference rules are iteratively applied to a knowledge base to extract more data until a goal is reached.

Definition: A mathematical transform that decomposes functions depending on space or time into functions depending on spatial or temporal frequency, often used in signal processing and analysis.

Definition: A form of many-valued logic or probabilistic logic; it deals with reasoning that is approximate rather than fixed and exact. Unlike traditional binary sets (where variables may take on true or false values), fuzzy logic variables may have a truth value that ranges in degree between 0 and 1. This allows for a more human-like way of thinking in machines, making it particularly useful in fields like control systems and decision-making where variables can range across a spectrum of possibilities rather than being strictly yes/no.