The field of machine learning can be thought of as the branch of AI (artificial intelligence) that is specifically aimed at building systems that learn from data without being explicitly programmed. A typical machine learning system is designed to uncover the deep relationships that are hidden in large data sets. The system learns these relationships, and uses them to generate non-obvious insights and actionable conclusions. The need for this capability is so great that it has generated a new buzzword, “Big Data.”
At Galois, we have extensive experience both in the application of the most effective machine learning techniques to a wide range of data sets and also in the design of new machine learning languages and frameworks that are especially suited to noisy and ambiguous data. In particular, we are working on a whole new class of “probabilistic programming languages” that promise to significantly enhance our abilities to reason about large data sets.
