The proposed model is based on the stochastic fractal search (SFS) – a powerful metaheuristic – and a nature-inspired algorithm that is claimed to solve complex optimization problems. The ...

Furthermore, gradient information is very much unavailable or is spurious. Stochastic optimization algorithms are even more suitable when it comes to coming up with the global optimum solution.

Within any application category or set of characteristics there are many optimization algorithms that are equivalently effective. Criteria for algorithm preference include robustness to surface ...

Unlike some optimization algorithms, differential evolution doesn't need to explicitly track the best solution found because the best solution will always be in the population. If you modify the basic ...

Continuously refining the model with each new observation, stochastic optimization algorithms have been enormously successful in scaling up machine learning approaches used in deep learning, ...

Jean-Jacques Forneron, Serena Ng, Estimation and Inference by Stochastic Optimization, AEA Papers and Proceedings, Vol. 111, PAPERS AND PROCEEDINGS OF THE One Hundred Thirty-Third Annual Meeting OF ...

We introduce and study three stochastic variance-reduced majorization-minimization (MM) algorithms, combining the general MM principle with new variance-reduced techniques. We study the almost surely ...

Advised by ECE professor Andrew Teel, Crisafulli studies stochastic approximation theory applied to hybrid dynamical systems. Although theoretical, the work can be applied to establish performance ...