CMSC421: Evolutionary Computation(100% Errorless solutions)

when to use Evolutionary Computation correct answers space to be search is large, and best solution is not necessarily required types of evolutionary computation correct answers genetic algorithms, evolution strategies, genetic programming basic idea of genetic algorithms correct answers create chromosomes that represent information about an individual, assess fitness, select most fit to reproduce into a new generation reproduction correct answers fitness based selection with replacement, multiple ways to choose the best recombination/crossover correct answers single point crossover, choose a point in 2 chromosomes, and then swap halves mutation correct answers point mutation, just mutate a single bit parameters of GAs correct answers population size n, probability of mutation p_m, probability of crossover p_c, maximum number of generations convergence correct answers we want best and average to converge later on, rather than earlier schema correct answers a template that matches chromosomes/strings, order o(h) = # of constant bits, distance d(h) = distance between first and last constant (exclude the first when counting) assessment of GAs correct answers can work very effectively, but computationally expensive and hard to represent some solutions as chromosomes evolution strategies correct answers developed for engineering optimization problems, a chromosome is now a vector of object parameters and a vector of strategy parameters, mostly mutations are done strategy parameters correct answers tells the likelihood of mutation for the corresponding position in the object parameters vector 1/5th rule correct answers if at least 1/5 of the time you are making positive changes, then increase the strategy parameter value genetic programming correct answers similar to GAs except it evolves trees, programs, networks, broadens the applicability of evolutionary computation