The presenters in this session will explore deep learning and how it can be relevant to actuaries. They will begin with an overview of concepts in neural network architecture and model training, then discuss an application to modeling shock lapses in post-level premium period term plans, including implementation details using R. Practical concerns such as how to interpret and communicate model results from deep learning models will be addressed. Presenters will also discuss potential applications of deep learning to other areas of actuarial science, underwriting and claims analytics.