paper-with-me

Papers

Variable Metric Evolution Strategies for High-dimensional Multi-Objective Optimization

2024-12-20 · Tobias Glasmachers

We design a class of variable metric evolution strategies well suited for high-dimensional problems. We target problems with many variables, not (necessarily) with many objectives. The construction combines two independent developments: efficient algorithms for scaling covariance matrix adaptation to high dimensions, and evolution strategies for multi-objective optimization. In order to design a specific instance of the class we first develop a (1+1) version of the limited memory matrix adaptation evolution strategy and then use an established standard construction to turn a population thereof into a state-of-the-art multi-objective optimizer with indicator-based selection. The method compares favorably to adaptation of the full covariance matrix.

📄 PDF Abstract BibTeX arXiv:2412.15647

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Challenges in High-dimensional Reinforcement Learning with Evolution Strategies

2018-06-04 · Nils Müller, Tobias Glasmachers

Evolution Strategies (ESs) have recently become popular for training deep neural networks, in particular on reinforcement learning tasks, a special form of controller design. Compared to classic problems in continuous di…

reinforcement-learningReinforcement LearningReinforcement Learning (RL)Vocal Bursts Intensity Prediction

Residual Learning Inspired Crossover Operator and Strategy Enhancements for Evolutionary Multitasking

2025-03-27 · Ruilin Wang, Xiang Feng, Huiqun Yu, Edmund M-K Lai

In evolutionary multitasking, strategies such as crossover operators and skill factor assignment are critical for effective knowledge transfer. Existing improvements to crossover operators primarily focus on low-dimensio…

Super-ResolutionTransfer Learning

An Invariant Information Geometric Method for High-Dimensional Online Optimization

2024-01-03 · Zhengfei Zhang, Yunyue Wei, Yanan Sui

Sample efficiency is crucial in optimization, particularly in black-box scenarios characterized by expensive evaluations and zeroth-order feedback. When computing resources are plentiful, Bayesian optimization is often f…

Bayesian OptimizationMuJoCo

Non-linear PCA via Evolution Strategies: a Novel Objective Function

2026-02-03 · Thomas Uriot, Elise Chung arxiv

Principal Component Analysis (PCA) is a powerful and popular dimensionality reduction technique. However, due to its linear nature, it often fails to capture the complex underlying structure of real-world data. While Ker…

Dimensionality Reduction

Guided Evolutionary Strategies: Escaping the curse of dimensionality in random search

2019-05-01 · ICLR 2019 5 · Niru Maheswaranathan, Luke Metz, George Tucker, Dami Choi 외

Many applications in machine learning require optimizing a function whose true gradient is unknown, but where surrogate gradient information (directions that may be correlated with, but not necessarily identical to, the …

Meta-LearningReinforcement Learning