OncoGuard-AI Synthetic Oncology Dataset

OncoGuard-AI Synthetic Oncology Dataset v2.1 is a reproducible synthetic dataset created for the development and academic evaluation of an oncology early-warning and clinical decision-support prototype. The dataset contains 14,998 records and 87 total columns: 74 engineered clinical features, 3 metadata columns, and 10 multiclass risk labels. The package includes the dataset CSV, reproducible data-generation source code, documentation, citation metadata, dependency versions, and separate licenses for data and code. The records are entirely synthetic and contain no real patient or personally identifiable information. This dataset is intended solely for research, education, thesis evaluation, and software demonstration. It has not been clinically validated and must not be used for diagnosis, treatment decisions, or production clinical care. CSV SHA-256: c1cfe9acbaf0e377a2a9dfe034f210857f8bda1964e5d6749c1df7b38de78300