Machine Learning is not magic. You can’t just throw the data through an algorithm and expect it to provide insights. You have to prepare the data and very often you have to tune the algorithm. Some algorithms - Neural Nets, Deep Learning, Support Vector Machines and Nearest Neighbour - are starting to dominate the field. A great deal of attention is often focused on the maths behind these, and it IS fascinating.
But you don’t have to understand the maths to be able to use these algorithms effectively. What you do need
to know is how they work because that is the information that allows you to tune them effectively. This talk will explain how they work from a non-mathematical standpoint.
