-
Notifications
You must be signed in to change notification settings - Fork 100
Home
ZOOpt is a python package for zeroth-order optimization, which is also called as derivative-free optimization and black-box optimization. It also covers algorithms under the names of evolutionary algorithms, metaheuristic algorithms, and direct search. They can deal with complex optimization problems such as functions that are not differentiable, not continuous, with many local optima, and even unknown but only testable.
ZOOpt is an optimization toolbox with three aims. First, ZOOpt collects theoretical grounded derivative-free optimization algorithms. Second, ZOOpt focuses on optimization problems in machine learning, e.g., high-dimensional, noisy, and large-scale optimizations. Third, ZOOpt tries to be a ready-to-use tool for accomplishing real-world machine learning tasks.
This documentation intends to explain what ZOOpt can do and how to use it.
Table of Content