GPR-132 Hyperparameter Optimization in Neural Network Using Binary Search Algorithm

Hyperparameter searching is a crucial process for every neural network training. However, this process is notably time-consuming due to the vast number of possible combinations and the influence these hyperparameters have on each other. The common approach is using grid search to exhaust all the options, which is computationally very expensive. In this research, we propose a new algorithm for this problem that is inspired by binary search and returns a significant improvement in time efficiency.