Fig. 1: Overview of the process-synergistic active learning framework.
From: A process-synergistic active learning framework for high-strength Al-Si alloys design
The framework comprises following main components. a Database construction through integration of experimental with literature-derived composition-process-property data, encompassing seven principal alloy elements beyond Al and Si, with PRs encoded via one-hot encoding. b Generation of process-specific alloy compositions latent space utilizing a c-WAE architecture. c Implementation of an ensemble surrogate model for performance prediction based on PRs and alloy compositions. d Systematic selection of top-ranked alloy compositions for experimental validation, followed by iterative database enrichment for model refinement until high-strength alloys are found.
