Datasets for Insights into prismatic loop formation in irradiated Fe-Cr alloys from hypothesis-driven active learning and causal analysis

Datasets for irradiated Fe-Cr alloys are collected from the experimental reports on dislocation loop type and dislocation density. We have constructed a data set from experimental literature containing 182 data points. To address such challenges to predict dislocation density, we have implemented a three-step ML approach as listed in the following: impute dataset to fill in the missing data to construct a predictive model using the RF regression algorithm. generate functionalized features and evaluate feature importance using the predictive model use the physics-based important functionalized features as hypotheses (physics-augmented GP models) in a hypothesis-driven active learning scheme to learn and predict dislocation density for all alloys.