Task: There are 5 projects. 5 tests are given in 5 different subjects. Each test has 5 questions. Users pass tests, and based on their answers, a table is formed in which 0 or 1 is indicated, i.e. correctly answered the question or not. Need to distribute users to projects based on their responses. In each project there can be 5 people. I do not understand how to carry out the crossing so that descendants are obtained, because in fact we only have fixed data, and we cannot change them, and create new participants.

There was an idea that you just need to take as a gene all the answers for one test, and already apply the GA to each test separately. Those. For example, for each project there is a definite assessment of how much you need to score for each test. And apply the HA to the test itself.

  • @rdorn thanks, now I understand - ATOM

1 answer 1

If the result of the algorithm should be the distribution of people according to the projects, then these distributions should be crossed in order to get the optimal distribution according to the given criteria. And the test results are needed to evaluate the distributions obtained during the GA operation.