Devoxx 2014 - Probably, Definitely, Maybe

Probably, Definitely, Maybe

We live in an uncertain world and we deal with this by considering possibilities and probabilities; from what movies you might be interested in to what diseases you might have given your symptoms. Without realising it people interact with systems that utilise probability on a daily basis. But what is probability, how to we measure and work with it? Given the wealth of big data collected how can we use that to predict the future? What if there is noise in the data? What if the data changes over time?

To answer these questions we start with the fundamentals of probability theory before looking Bayesian systems (frequently a key part of recommendation and expert systems), Markov chains (useful in predicting human behaviour), and dynamic Bayesian networks (such as Kalman filters and hidden Markov chains which are often used in real-time systems like auto-pilot navigation). Along the way we'll try to demonstrate the mathematics using simple Java/Groovy examples. No knowledge of advanced mathematics is required!

Source: DEVOXX