Glossary of Key Terms
Glossary of Key Terms
| Term | Definition |
| Allele | The value a gene takes for a particular chromosome. |
| Baldwinian Model | A model of lifetime adaptation where chromosomes encode a tendency for learning beneficial behaviors; improved fitness is assigned but the acquired trait itself is not passed to the next generation. |
| Building Block Hypothesis | The idea that GAs succeed by identifying and recombining low order, low defining length schemata with above-average fitness. |
| Chromosomes | A single candidate solution to a given problem. |
| Classifier Systems | A form of genetics-based machine learning (GBML) system. |
| Crossover | A genetic operator where more than one parent is selected and one or more offspring are produced using the genetic material of the parents. It is analogous to reproduction. |
| Crowding | A phenomenon where a highly fit chromosome reproduces excessively, causing the population to fill with similar solutions and reducing diversity. |
| Decoding | The process of transforming a solution from the genotype (computation) space to the phenotype (real-world) space. |
| Elitism | A survivor selection policy where the current fittest member of the population is guaranteed to be propagated to the next generation. |
| Encoding | The process of transforming a solution from the phenotype (real-world) space to the genotype (computation) space. |
| Evolutionary Computation | A larger branch of computation of which Genetic Algorithms are a subset, based on concepts of natural selection and genetics. |
| Fitness Function | A function that takes a solution as input and produces the suitability of the solution as the output. |
| Gene | One element position of a chromosome. |
| Genetic Algorithm (GA) | A search-based optimization technique based on the principles of Genetics and Natural Selection, used to find optimal or near-optimal solutions. |
| Genotype | The population represented in the computation space, in a way that can be easily manipulated by a computing system. |
| Lamarckian Model | A model of lifetime adaptation which states that traits an individual acquires in its lifetime can be passed on to its offspring. |
| Mutation | A small random tweak in a chromosome to get a new solution, used to maintain and introduce diversity in the population. |
| No Free Lunch (NFL) Theorem | A theorem stating that when averaged over all possible problems, all non-revisiting black box algorithms exhibit the same performance. |
| NP-Hard | A class of computational problems for which even the most powerful computing systems take a very long time to solve. |
| Optimization | The process of making something better by finding the values of inputs that result in the best possible output values (e.g., maximizing or minimizing an objective function). |
| Phenotype | The population in the actual real-world solution space, where solutions are represented as they are in real-world situations. |
| Population | A subset of all possible encoded solutions to a problem in the current generation; a set of chromosomes. |
| Premature Convergence | An undesirable condition where one extremely fit solution takes over the entire population, leading to a loss of diversity and causing the GA to get stuck. |
| Schema | A template string over the alphabet {0,1,*} that represents a hyperplane in the solution search space. |
| Schema Theorem | A theorem stating that schema with above-average fitness, short defining length, and lower order are more likely to survive crossover and mutation. |
| Search Space | The set of all possible solutions or values which the inputs can take for a given problem. |
| Selection | The process of choosing parents which mate and recombine to create offspring for the next generation. |