HALICON 2017

Robots may never completely replace doctors and nurses, but machine learning is nevertheless transforming the healthcare industry by improving diagnostics and predicting outcomes. But while healthcare is all about trust, machine learning models often are black boxes – which makes it inherently difficult to trust them. How do patients and doctors know that they can trust predictions from a model that they cannot understand? 

Transparency in machine learning models is critical in high stakes decisions, like those made every day in healthcare. This presentation will cover all important phases of data science including the ingestion of data, the cleaning and blending of data sets, finding the best models for the task, validating the quality of those models, and finally putting them into production. We will demonstrate, how analysts can achieve highly accurate models in a way that allows others to understand the results of their work.  In particular, we will create a predictive model and validate its correctness without writing a single line of code and in an hour instead of weeks or months.