Improving Evolutionary Strategies with Generative Neural Networks
arXiv:1901.11271
Abstract
Evolutionary Strategies (ES) are a popular family of black-box zeroth-order optimization algorithms which rely on search distributions to efficiently optimize a large variety of objective functions. This paper investigates the potential benefits of using highly flexible search distributions in classical ES algorithms, in contrast to standard ones (typically Gaussians). We model such distributions with Generative Neural Networks (GNNs) and introduce a new training algorithm that leverages their expressiveness to accelerate the ES procedure. We show that this tailored algorithm can readily incorporate existing ES algorithms, and outperforms the state-of-the-art on diverse objective functions.
References in corpus (9)
- NICE: Non-linear Independent Components Estimation
- Evolution Strategies as a Scalable Alternative to Reinforcement Learning
- Density estimation using Real NVP
- CMA-ES for Hyperparameter Optimization of Deep Neural Networks
- Counterfactual Risk Minimization: Learning from Logged Bandit Feedback
- Backpropagation through the Void: Optimizing control variates for black-box gradient estimation
- High-Dimensional Probability Estimation with Deep Density Models
- A comparative study of counterfactual estimators
- Neural Generative Models for Global Optimization with Gradients