Not quite unreasonable effectiveness of machine learning algorithms

Egor Illarionov; Roman Khudorozhkov

Not quite unreasonable effectiveness of machine learning algorithms

Computer Vision and Pattern Recognition 2018-07-25 v1 Machine Learning

Authors: Egor Illarionov , Roman Khudorozhkov

Abstract

State-of-the-art machine learning algorithms demonstrate close to absolute performance in selected challenges. We provide arguments that the reason can be in low variability of the samples and high effectiveness in learning typical patterns. Due to this fact, standard performance metrics do not reveal model capacity and new metrics are required for the better understanding of state-of-the-art.

Keywords

machine learning algorithm selection machine learning theory

Cite

@article{arxiv.1804.02543,
  title  = {Not quite unreasonable effectiveness of machine learning algorithms},
  author = {Egor Illarionov and Roman Khudorozhkov},
  journal= {arXiv preprint arXiv:1804.02543},
  year   = {2018}
}

Related papers

View all related →