evolutionary computation
In plain wordsProblem-solving by simulated evolution, much as breeders improve crops: keep many candidate answers, let the better ones produce slightly altered offspring, drop the worse, and repeat for many rounds.
The research field, and the family of algorithms it has produced, that tackles optimization and design problems by simulating Darwinian variation and selection over many generations. An evolutionary algorithm maintains a population of candidate solutions, scores each with a fitness functionfitness function, n.: In evolution-inspired computing, the scoring rule that says how good each candidate solution is, which decides which candidates survive and get to 'reproduce'., selects fitter ones as parents and produces offspring by variation — mutation and, in many variants, recombination — repeating until a budget or quality target is reached. The historical branches differ mainly in representation and emphasis: genetic algorithmsgenetic algorithm, n.: A problem-solving method that copies natural evolution: many candidate solutions compete, the best are 'bred' by mixing and randomly tweaking them, and over many generations the solutions improve. (strings and crossover), evolution strategiesevolution strategies, n.: A family of evolution-inspired methods for tuning a list of numbers: try many random variations of the current best settings, keep those that work better, and adjust how big the variations are as you go. (real-valued vectors with self-adapting mutation), evolutionary programming (originally finite-state machines) and genetic programminggenetic programming, n.: Using simulated evolution to write computer programs: random programs compete at a task, and the best are combined and mutated over many generations until one does the task well. (programs). Later additions include differential evolution, estimation-of-distribution algorithms, multi-objective methods such as NSGA-II, neuroevolutionneuroevolution, n.: Training or designing neural networks by simulated evolution — breeding and mutating whole networks and keeping the best performers — instead of, or alongside, the usual gradient-based training. and quality diversityquality diversity, n.: A family of evolution-inspired search methods that aim to collect many different good solutions — for example, many distinct ways for a robot to walk — rather than a single best one. algorithms. Since 2022, beginning with OpenAI's Evolution through Large Models, systems such as FunSearchFunSearch, prop. n.: A Google DeepMind method that discovered new mathematical constructions by having a language model repeatedly rewrite short computer programs, keeping the versions that scored best. and AlphaEvolveAlphaEvolve, prop. n.: A Google DeepMind system that combines AI language models with an evolution-like search to invent and improve computer programs, finding new results in mathematics and computing. have used language modelslarge language model, n.: A computer program trained on enormous amounts of text to predict the words that come next. At vast scale, that simple skill lets it chat, answer questions, translate, summarize and write code. as mutation operators that rewrite candidate programs, renewing mainstream interest in the field.
Origin
Computer simulations of evolution began with Nils Aall Barricelli (1953–1954) and Alex Fraser (1957). Independent schools followed in the 1960s: Lawrence Fogel's evolutionary programming in the United States, Ingo Rechenberg and Hans-Paul Schwefel's evolution strategies in Germany, and John Holland's genetic algorithms. The umbrella term spread in the early 1990s, marked by the journal Evolutionary Computation (1993).
Related
- genetic algorithmgenetic algorithm, n.: A problem-solving method that copies natural evolution: many candidate solutions compete, the best are 'bred' by mixing and randomly tweaking them, and over many generations the solutions improve.
- evolution strategiesevolution strategies, n.: A family of evolution-inspired methods for tuning a list of numbers: try many random variations of the current best settings, keep those that work better, and adjust how big the variations are as you go.
- genetic programminggenetic programming, n.: Using simulated evolution to write computer programs: random programs compete at a task, and the best are combined and mutated over many generations until one does the task well.
- neuroevolutionneuroevolution, n.: Training or designing neural networks by simulated evolution — breeding and mutating whole networks and keeping the best performers — instead of, or alongside, the usual gradient-based training.
- quality diversityquality diversity, n.: A family of evolution-inspired search methods that aim to collect many different good solutions — for example, many distinct ways for a robot to walk — rather than a single best one.
- swarm intelligenceswarm intelligence, n.: The teamwork that appears when crowds of simple creatures or machines, such as ants, bees or small robots, each respond only to what is nearby, with nobody in charge, yet together solve problems.
- AlphaEvolveAlphaEvolve, prop. n.: A Google DeepMind system that combines AI language models with an evolution-like search to invent and improve computer programs, finding new results in mathematics and computing.
- metaheuristicmetaheuristic, n.: A general-purpose strategy for finding good answers to hard problems, such as imitating evolution or cooling metal, that can be adapted to many different problems instead of being designed for just one.
Sources
- Fogel, L. J., Owens, A. J. & Walsh, M. J. (1966). Artificial Intelligence through Simulated Evolution. Wiley.
- Eiben, A. E. & Smith, J. E. (2015). Introduction to Evolutionary Computing, 2nd ed. Springer.
Mentioned in
genetic algorithmgenetic algorithm, n.: A problem-solving method that copies natural evolution: many candidate solutions compete, the best are 'bred' by mixing and randomly tweaking them, and over many generations the solutions improve., swarm intelligenceswarm intelligence, n.: The teamwork that appears when crowds of simple creatures or machines, such as ants, bees or small robots, each respond only to what is nearby, with nobody in charge, yet together solve problems., ant colony optimizationant colony optimization, n.: A method for finding good routes and schedules inspired by how ants find short paths to food: simulated ants leave 'scent' on good paths, which attracts other ants, so the best routes gradually stand out., artificial lifeartificial life, n.: The study of life by building lifelike systems — in computer simulations, robots or chemistry — to learn how living things organize themselves, reproduce and evolve, and what other kinds of life could exist., computational intelligencecomputational intelligence, n.: A family of AI methods inspired by nature, such as neural networks, fuzzy logic and evolutionary algorithms, that learn or adapt from numbers rather than following explicit logical rules., evolution strategiesevolution strategies, n.: A family of evolution-inspired methods for tuning a list of numbers: try many random variations of the current best settings, keep those that work better, and adjust how big the variations are as you go., fitness functionfitness function, n.: In evolution-inspired computing, the scoring rule that says how good each candidate solution is, which decides which candidates survive and get to 'reproduce'., FunSearchFunSearch, prop. n.: A Google DeepMind method that discovered new mathematical constructions by having a language model repeatedly rewrite short computer programs, keeping the versions that scored best., genetic programminggenetic programming, n.: Using simulated evolution to write computer programs: random programs compete at a task, and the best are combined and mutated over many generations until one does the task well., John HollandJohn Holland, prop. n.: An American scientist who invented genetic algorithms, which solve problems by breeding and mutating populations of candidate answers, and helped found the science of complex systems., metaheuristicmetaheuristic, n.: A general-purpose strategy for finding good answers to hard problems, such as imitating evolution or cooling metal, that can be adapted to many different problems instead of being designed for just one., neural architecture searchneural architecture search, n.: Using an automated search, rather than human trial and error, to design the structure of a neural network: which layers to use, how many, and how to connect them., neuroevolutionneuroevolution, n.: Training or designing neural networks by simulated evolution — breeding and mutating whole networks and keeping the best performers — instead of, or alongside, the usual gradient-based training., particle swarm optimizationparticle swarm optimization, n.: An optimization method inspired by flocking birds: candidate solutions 'fly' through the space of possibilities, each pulled toward the best spot it has found and the best spot its neighbors have found., quality diversityquality diversity, n.: A family of evolution-inspired search methods that aim to collect many different good solutions — for example, many distinct ways for a robot to walk — rather than a single best one., soft computingsoft computing, n.: A group of computing methods, including fuzzy logic, neural networks and evolutionary algorithms, that accept imprecise, approximate answers in exchange for being practical on messy real-world problems., swarm roboticsswarm robotics, n.: Coordinating large groups of simple robots that follow local rules and talk only to their neighbors, so useful group behavior emerges without a central controller, as in ant colonies or bird flocks., symbolic regressionsymbolic regression, n.: Getting a computer to discover a readable mathematical formula, such as a law of physics, that fits a set of measurements, instead of only fitting numbers into a formula chosen in advance.