Black Box and Data-Driven Computation
Published in Book Black Box Optimization, Machine Learning and No-Free Lunch Theorems, 2020
Recommended citation: Rong Jin, Weili Wu, My T. Thai, and Ding-Zhu Du. Book Black Box Optimization, Machine Learning and No-Free Lunch Theorems.
Abstract
The black box was an important tool in study of computational complexity theory and is used for establishing the hardness of problems. The data-driven makes it be utilized as a tool for proving solutions for some computational problems. In this note, we present several observations for this new role of reduction.