Predicting Drug Approvals: The Novartis Data Science and Artificial Intelligence Challenge

We describe a novel collaboration between academia and industry, an in-house data science artificial intelligence challenge held by Novartis to develop machine learning models for predicting drug development outcomes, building upon research at MIT using from Informa® as the starting point. Over 50 cross-functional teams 25 offices around world participated in challenge. The domain expertise of these researchers was leveraged create predictive with greater sophistication, two developed that outperformed baseline model through state-of-the-art algorithms use newly incorporated features data. In addition validating variables shown be associated approval earlier study, also provided new insights into drivers success failure.