Published June 7, 2026 | Version 0.2.0

wafer-proc-sim: Data-quality-aware heteroscedastic GP surrogate for 4H-SiC blade dicing

Authors/Creators

  • 1. Keio University / Leibniz University Hannover

Description

Open-source framework for 4H-SiC blade dicing process optimization combining physics-based simulation, heteroscedastic Gaussian process surrogates, Ensemble Kalman Filter state estimation, and Bayesian optimization. Core paper path is reproducible via scripts/reproduce_paper.sh with CI-regression-tested LOO metrics.

Files

keisuke58/wafer-proc-sim-v0.3.0.zip

Files (26.7 MB)

Name Size Download all
md5:b555cc714b93beb785c6709e8d80dd08
26.7 MB Preview Download

Additional details

Related works