Researchers detail TrojAI, a framework for hardening AI models against adversarial attacks Source: venturebeat.com One way to test machine learning models
