Meta-Heuristics Flashcards
What are Genetic Algorithms?
Genetic algorithms are a type of evolutionary algorithm that mimic natural selection processes, where a population of candidate solutions to an optimization problem evolves toward better solutions.
What are the main components of a Genetic Algorithm?
The main components include a population of chromosomes (possible solutions), genetic operators such as selection, crossover, and mutation, and a fitness function to evaluate solutions.
How does selection work in Genetic Algorithms?
Selection involves choosing fitter chromosomes from the population for reproduction based on their fitness scores, influencing the likelihood that certain features will pass to the next generation.
What is Simulated Annealing?
Simulated Annealing is a probabilistic technique for approximating the global optimum of a given function. It is inspired by the process of annealing in metallurgy where material is heated and then slowly cooled to increase the size of its crystals and reduce defects.
What is crossover in the context of Genetic Algorithms?
Crossover is a genetic operator used to combine the genetic information of two parents to generate new offspring, contributing to genetic variety in the population.
What role does mutation play in Genetic Algorithms?
Mutation introduces random alterations to individual genes, providing genetic diversity and helping the algorithm to avoid local minima by exploring new areas in the solution space.
How does Simulated Annealing work?
The algorithm starts with an initial solution and explores its neighborhood. Moves to solutions with lower cost are always accepted, and moves to higher cost solutions are accepted with a probability that decreases with the temperature—a control parameter that gradually decreases.
What is the role of temperature in Simulated Annealing?
Temperature in Simulated Annealing controls the probability of accepting worse solutions as the algorithm proceeds, helping it to escape local minima and explore globally for better solutions.
What is a cooling schedule in Simulated Annealing?
The cooling schedule is a plan that defines how the temperature should decrease in the course of the algorithm. Key parameters include initial temperature, cooling rate, and the number of iterations at each temperature.
How are new solutions generated in Simulated Annealing?
New solutions are typically generated by making small, random changes to the current solution, ensuring exploration of the solution space. The acceptance of new solutions depends on whether they improve the solution or not, and according to the temperature-driven probability for worse solutions.
