What question did this study set out to answer?

This research evaluates the effectiveness of the EGO-Optimizer in large-scale optimization tasks.

February 16, 2026Open Access

Scalability Analysis of Entropy-Guided Optimization in High Dimensions

Key Points

This research evaluates the effectiveness of the EGO-Optimizer in large-scale optimization tasks.
Introduced a feedback loop based on Shannon-entropy to enhance population diversity.
Benchmarked the optimizer on landscapes of 30D, 50D, and 100D dimensions.
Compared performance against standard metaheuristics like DE, PSO, and GA across various benchmarks.
Demonstrated high scalability in high-dimensional tasks compared to traditional methods.
Adapted exploration-exploitation trade-off improved performance and convergence rates.

Abstract

This paper presents the EGO-Optimizer, a Python library implementing Entropy-Guided Optimization (EGO) for large-scale continuous optimization tasks. The study introduces a Shannon-entropy feedback loop to quantify population diversity in real-time, allowing for adaptive control of the exploration-exploitation trade-off. Key Contributions: Entropy-Driven Mutation: Replaces fixed scaling with a real-time entropy-normalized factor to prevent premature convergence. High-Dimensional Scalability: Benchmarked on 30D, 50D, and 100D landscapes against standard metaheuristics (DE, PSO, GA). Performance Validation: Evaluated across 67 CEC-style benchmark instances (unimodal, multimodal, and composition functions). Source Code: The full implementation, documentation, and usage examples are available on GitHub:https://github.com/Draster2k/ego

Read Full Paperexternally

اسأل الذكاء الاصطناعي

Bookmark

View Full Paper