120 likes | 207 Views
Valery Frolov. Final report. Content of the talk. The algorithm Fitness function Crossover Mutation Elite individuals Reverse mutations Some statistics Run examples. The algorithm. The most standard evolution algorithm While (condition isn’t met) Select fittest individuals
E N D
Valery Frolov Final report
Content of the talk • The algorithm • Fitness function • Crossover • Mutation • Elite individuals • Reverse mutations • Some statistics • Run examples
The algorithm • The most standard evolution algorithm • While (condition isn’t met) • Select fittest individuals • Crossover • Mutation • Update population • The condition • Simulation time
Fitness function • The parameters we have • Best laptime • Top speed • Distance raced • Damage done • The function is :
Crossover function • Ideal crossover function forour problem • Learns the relation between different parameters • Performs the crossover accordingly • My crossover • Interleaves two parents with equal probability • Creates one sibling
Mutation function • Each one of 22 parameters is mutated with a probability of 0.1 • Therefore, each individual is mutated with a high probability
Updating the population • Each new population consists of • Elite individuals from previous population • Individuals created by crossover • Several random individuals
Elite individuals • Each population consists of K individuals • Best M individuals are copied to the next population. • M is 1/3 of population size • Crossovers are done with at least one elite parent
Reversed mutations • Problem : • Mutations can make things worse • Elite individual that survived many generation can be removed because of one bad mutation • Our solution : • Reverse previous mutation if condition is satisfied
Size of population is 16 • Time per each individual is 4450 ticks • Individuals get bonus points for surviving • Individuals get penalty for non-completed laps
Some statistics • In most cases starting population consists of pair of “good” and many “bad” individuals • At least half of final generation are “very good” individuals