AI 용어집
인공지능 완전 사전
Crossover Operator
Genetic operation that combines the genetic material of two parents to create one or more offspring, promoting the propagation of advantageous characteristics.
Mutation Operator
Random modification of an individual's genome that introduces genetic diversity into the population, allowing to avoid local optima.
Fitness Function
Evaluation function that quantifies the quality or adaptability of a candidate solution with respect to the objective to optimize.
Ant Colony Algorithms
Metaheuristic inspired by the behavior of ants that find optimal paths through the deposition and tracking of pheromones.
Evolution Strategies
Evolutionary paradigm using self-adaptive mutation as the main operator, particularly effective for continuous optimization.
Memetic Algorithms
Hybridization of evolutionary algorithms with local search techniques to accelerate convergence towards optimal solutions.
Evolutionary Niching
Technique preserving solution diversity by maintaining subpopulations in different ecological niches of the search space.
Adaptive Landscape
Multidimensional representation of the fitness function where each point corresponds to a solution and its height to its evaluative quality.
Selective Pressure
Intensity with which natural selection favors the most adapted individuals, influencing the algorithm's convergence speed.
Genetic Drift
Stochastic phenomenon where allele frequencies change randomly in a population, potentially leading to a loss of genetic diversity.
Gray Code
Binary coding system where two successive values differ by only one bit, optimizing mutations in genetic algorithms.
Multi-objective Evolutionary Algorithm
Extension of evolutionary algorithms that simultaneously optimizes several conflicting objectives to find a set of Pareto-optimal solutions.
Holland's Schema
Theory describing how building blocks (schemas) of partial solutions are combined and propagated through generations in genetic algorithms.