How to make decisions with very little data: a guide to Bayesian Inference
11-Oct-2018
AI and Machine Learning Theatre
Despite the data revolution, most corporations only have very little data. In fact, nobody has "big data", because if you do, then you can always slice your dataset across a various groups repeatedly, and now you are back in the "small data" regime. How can one efficiently leverage small amounts of data to make decisions in the face of uncertainty? This talk will introduce Bayesian Inference, which is an under-appreciated but incredibly effective methodology for leveraging small data sets for decision making. Bayesian inference was instrumental in cracking the Enigma code in World War II, searching for missing submarines, and a numerous other situations where critical decision making was necessary, the stakes were high and the data scarce. The talk will introduce Bayesian Inference in way that is accessible to a wide audience and point the listeners to a list of tools and resources they can use immediately in their own business context.