Discrete Natural Evolution Strategies

Natural evolution strategies are a class of approximate-gradient black-box optimizers that have been successfully used for continuous parameter spaces. In this paper, we derive NES algorithms for discrete parameter spaces and demonstrate their effectiveness in tasks involving discrete parameters.

cs.NE math.OC Evolutionary Algorithms and Applications Computer Science Mathematics

View graph Source

Paper

References (6)

01Few-Shot Bayesian Imitation Learning with Logical Program PoliciesTom Silver, Kelsey R. Allen, Alexander K. Lew et al.2019 · AAAI Conference on Artificial Intelligence · 60 citations In Library

02Evolution Strategies as a Scalable Alternative to Reinforcement LearningTim Salimans, Jonathan Ho, Xi Chen et al.2017 · arXiv.org · 1.8k citations In Library

03New Insights and Perspectives on the Natural Gradient MethodJames Martens2014 · Journal of machine learning research · 836 citations In Library

04A method for stochastic optimization2017

05Fisher information matrix for Gaussian and categorical distributions2012 · Fisher_inf.pdf

06Program synthesis by sketchingRastislav Bodík, Armando Solar-Lezama2008 · 557 citations

Similar papers

Natural Evolution StrategiesDaan Wierstra, T. Schaul, T. Glasmachers et al.2011 · Journal of machine learning research · 34 citations In Library

Distance-weighted Exponential Natural Evolution Strategy for Implicitly Constrained Black-Box Function OptimizationM. Nomura, N. Sakai, Nobusumi Fukushima et al.2021 · IEEE Congress on Evolutionary Computation · 11 citations In Library

Natural Evolutionary Strategies for Variational Quantum ComputationA. Anand, M. Degroote, Alán Aspuru-Guzik2020 · Machine Learning: Science and Technology · 45 citations In Library

A Linear Time Natural Evolution Strategy for Non-Separable FunctionsYi Sun, Faustino J. Gomez, T. Schaul et al.2011 · Annual Conference on Genetic and Evolutionary Computation · 36 citations In Library

Towards a Principled Learning Rate Adaptation for Natural Evolution StrategiesMasahiro Nomura, Isao Ono2021 · arXiv (Cornell University) In Library

Learning Discrete Structured Variational Auto-Encoder using Natural Evolution StrategiesAlon Berliner, Guy Rotman, Yossi Adi et al.2022 · International Conference on Learning Representations · 6 citations In Library

CoNES: Convex Natural Evolutionary StrategiesSushant Veer, Anirudha Majumdar2020 · arXiv (Cornell University) In Library

Natural evolution strategies and variational Monte CarloTianchen Zhao, Giuseppe Carleo, J. Stokes et al.2020 · Machine Learning: Science and Technology · 26 citations In Library

Natural Evolution Strategy for Mixed-Integer Black-Box OptimizationKoki Ikeda, Isao Ono2023 · Proceedings of the Genetic and Evolutionary Computation Conference In Library

Natural Evolution Strategy for Mixed-Integer Black-Box OptimizationKoki Ikeda, I. Ono2023 · Annual Conference on Genetic and Evolutionary Computation · 7 citations In Library

Source of record

arXiv

44B

Welcome.

This panel sticks with you. Pick anything below and it opens right beside it, so you can dig through 60,000-plus records without ever losing your spot here.

Travel 44B

LibraryPapers, policy, standards, statute LabsEvery organization building AI ModelsIntelligence, price, and speed BenchmarksThe evaluation catalog IncidentsAI failures and harms SearchOne field across everything DashboardThe state of AI accountability in NY 44B RegistryThe Art. 44-B compliance portal