Transparent Antenna Topology Six-Feature Dataset

Dataset supporting the study “Explainable Machine Learning for Transparent Antenna Topology: Interpreting Geometry–Radio-Frequency Performance Relationships.” This dataset contains 351 transparent antenna topology records used in the explainable machine-learning analysis. Each record contains six structural descriptors derived from 16×16 binary topology representations: Metal_Ratio, Connected_Components, Hole_Count, Horizontal_Symmetry, Vertical_Symmetry, and Edge_Occupancy, together with the associated RF Score used as the response variable. The dataset supports the Random Forest regression, out-of-fold Shapley Additive exPlanations (SHAP), principal component analysis (PCA), and descriptive decile analyses reported in the associated manuscript. The electromagnetic evidence underlying the study was obtained from CST Studio Suite simulations. This dataset is provided to support reproducibility and transparent reuse of the reported computational analysis.