Trained Surrogate Models for Predictive–Inverse Design of Cellular Steel Beams

Description

This dataset provides the trained surrogate-model assets developed for the study “An Intelligent Predictive–Inverse Design Framework for Cellular Steel Beams with Integrated Robustness, Sustainability Metrics, and Automated CAD Generation.” The files include the final monotonic LightGBM resistance predictor, the fitted input scaler required for inference, the trained NODE model weights, and the corresponding NODE configuration file. These assets enable reproduction of the forward resistance predictions used within the inverse-design workflow for laterally restrained cellular steel beams.

Categories

Licence