Director of Research at Google Inc., Peter Norvig is a reference in terms of Machine Learning. He’s always looking at the world to re-build it, playing with programs and applications that nowadays have a dominant role in our world. “The world is made of lines” and we can draw anything with it, manually or numerically.
Peter Norvig’s talk is about deploying machine learning applications in the Enterprise.
More companies are taking advantage of the opportunity to harness large amounts of data to make more accurate predictions, and sometimes to invent entire new classes of applications that couldn’t be done before. In this talk, Peter Norvig uses the exemple of Google’s photo reconnaissance or its Automatic spell-checker. “21 lines of code can now be enough to build a whole program which self-feeds”.
But like any engineering discipline, machine learning brings with it some potential pitfalls that practitioners must learn to avoid.