Agentic data science with the Data Agent Kit
cloud.google.com·7h ago
AI Research
arXiv CS.AI·Aug 26·5 min read
Coronavirus Optimization Algorithm: A Success-History Adaptive Evolutionary Framework with Archive-Assisted Search and Stagnation Recovery for Global Optimization
TL;DR
The Coronavirus Optimization Algorithm (COA) is a novel evolutionary optimizer inspired by coronavirus mechanisms, achieving superior performance on global optimization tasks. It outperformed 15 competitive optimizers across various dimensions, particularly excelling in composition functions.
✦ Why It Matters
Consider implementing COA for your optimization problems to leverage its superior performance on complex functions.
Key Takeaways
- Explore COA for box-constrained optimization tasks.
- Utilize elite-guided attraction for improved search efficiency.
- Evaluate performance on CEC 2017 benchmark functions.
- Investigate COA's limitations on hybrid functions.
How It Works
COA mimics certain mechanisms of the coronavirus, such as elite-guided attraction and adaptive parameter variation, to enhance its search capabilities. It uses a combination of opposition-based initialization and current-to-pbest mutation to generate trial vectors, while an external archive helps retain successful solutions, allowing for effective stagnation recovery and population management.