For these tests feed forward ANNs where used. They where all initialized with random weights at the beginning of each run. One population consists of 21 individuals. From these individuals 6 are choosen to participate at a match. 3 in the right and tree in the left team. They play a match of 200 steps. After the match the following parameters are collected for each player to generate a fitness valule.
|kickCount||The numper of ball kicks|
|otherGoalsCount||The number of goals a player scored|
|ownGoalsCount||The number of own goals a player scored|
The fitness is then calculated by the folowing formula
f = kickCount + 100 * otherGoalsCount - 100 * ownGoalsCount
After 100 matches from the fittest individuals a new generation is generated by the following rules.
Then a new simulation cycle is started with this new population.
Goal of this testrun was to determine the optimal value for mutation rate ('mutationRate').
The resultsOMR1 data set contains about 100 simulation runs which contain diffrent values of mutationRate. Each of these runs contain multiple observaions of fitness, goals scored ('otherGoalCount') and own goals scored ('ownGoalCount'). The frequency of observations for diffent values of mutationRate can be seen in the following diagram.
The fitness for these values can be seen in the following diagram.
A mutation rate of 10000(per million) seems to result in a higher fitness value than smaller and bigger values.
The otherGoalsCount for diffrent mutationRate values can be seen in the following diagram.
A mutation rate of 10000(per million) seems to result in more goals scored than smaller and bigger values.
The ownGoalsCount for diffrent mutationRate values can be seen in the following diagram.
The number of own goals scored seems to be independend of of the mutation rate. A higher dependency from the mutation rate could brobably be seen in longer runs. Simulations where always stopped after 30000 matches. The ownGoalCount is smaller than the otherGoalCount for all mutation rates.