The speaker will present a rudimentary mathematical perspective of machine learning. As an example, he will explain how linear algebra and a bit of differential geometry can be applied under the framework of machine learning to interpret user ratings and make recommendations.