ASM-EA: Reproducibility Artifact for "A Surrogate-Assisted Multi-mode Algorithm with Constraint Boundary Awareness for Tightly Constrained Expensive Optimization"
Year: 2026
Type: software
Abstract: Reproducibility artifact for ASM-EA, a surrogate-assisted multi-mode evolutionary algorithm with constraint boundary awareness for tightly constrained expensive optimization. The archive contains everything needed to rebuild the experimental environment and reproduce all reported results. Contents code/ — full source code: ASM-EA and baselines (SACCDE, GLoSADE, DE), Kriging surrogate with uncertainty, benchmark suites (CEC2006, CEC2010, CEC2017, engineering benchmarks, IEEE 30-bus carbon-constrained ED), and stage-by-stage experiment runners (baseline, full comparison, ablation, sensitivity, case studies, publication figures). environment/ — exact environment specification: pinned dependency lock file (requirements.lock.txt), requirements.txt, Conda environment.yml, and setup notes (Python 3.9). data/ — re-run experiment results backing the manuscript tables and figures. figures/ — publication figures in vector PDF and 600-DPI PNG. README.md — archive overview and reproduction instructions. Requirements: Python 3.9 with the pinned dependencies in environment/requirements.lock.txt (NumPy 2.0, SciPy, scikit-learn, matplotlib, pandas, joblib). CPU only; no GPU required.
Author: Wu, Tianyong
Institution: Hubei Polytechnic University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7220729372