Open vivekaxl opened 7 years ago
i think i can speed up part 3.
take a random config. pass it down the branch that goes to the best leaf. mutate that instance according to the nodes in that branch. evaluate that "best" mutant.
note that in your approach, step3 requires nearly 3.9 miilion evals for sql at each step. in mine, it requires only one
you need a name for your new method
NAIR: rating acquired incremental reasons
In our previous paper, we used a rank based progressive sampling technique to optimize for software systems. However progressive sampling uses a validation set to perform iterative sampling and that adds additional cost to the sampling process. Caution: none of the previous papers including our paper does not talk about the validation set.
This new technique is does not use any validation technique and hence requires far fewer evaluations hence lowering the cost of optimization by a factor of [2.28-46.45].
This technique is inspired by both active learning and rank based sampling. The process starts by