# What are Genetic Algorithms?

Genetic algorithms are mathematical structures using the procedure of genetic inheritance. They have been successfully used to a broad variety of analytic issues. Data mining can connect human understanding with automatic analysis of information to find a design or key relationships.

Given a large database represented over several variables, the objective is to effectively find the most interesting design in the database. Genetic algorithms have been used to recognize interesting designs in some software. They generally are used in data mining to enhance the execution of other algorithms, such as decision tree algorithms, another association rule.

Genetic algorithms needed a specific data structure. They work on a population with characteristics defined in categorical structure. The analogy with genetics is that the population (genes) includes characteristics. There is a method to implement genetic algorithms is to use operators (reproduction, crossover, selection) with the feature of mutation to improve the generation of probably better combinations.

The genetic algorithm procedure is as follows −

• It can randomly choose parents.

• It is used to recreate through the crossover.

• Reproduction is selecting which single entities will handle it. In other terms, some objective services or selection features are required to determine survival. Crossover describes changes in the future production of entities.

• It can choose survivors for the next generation through a fitness service.

• The mutation is the service by which randomly chosen attributes of randomly selected entities in the following operations are transformed.

• It can repeat until either a given fitness level is achieved, or the present number of iteration has arrived.

• Genetic algorithm parameters contain population size, crossover rate, and mutation rate.

The advantage of the Genetic algorithm is as follows −

• Genetic algorithms are very accessible to create and validate which creates them highly attractive if used.

• The algorithm is parallel, defining that it can be used to high populations efficiently. The algorithm is also effective in that if it starts with a poor original solution, it can promptly progress to the best solutions.

• The use of mutation creates the method adequate for recognizing global optima even in very nonlinear problem rules. The method does not need knowledge about the distribution of the information.

• Genetic algorithms needed mapping data sets to from where attributes have discrete values for the genetic algorithm to work with. This is generally possible but can lose a big deal of detailed data when dealing with continuous variables.

• It is used to code the information into categorical form can involuntarily lead to biases in the record.

• There are also check on the size of a data set that can be considered with genetic algorithms.

• For very huge data sets, sampling will be important, which leads to multiple results across several runs over the equivalent data set.

Updated on: 22-Nov-2021

276 Views