- Research Program Mentor
PhD Doctor of Philosophy candidate
Foundations of data science, machine learning, mathematics, probability, statistics
Looking inside the black box: why machine learning works
When we read about machine learning in media, it's often presented as a black box; that is, data goes in, the computer does some magic, and out comes a recommendation for which socks you will find most comfortable. While it is true that machine learning algorithms must be complex to solve complex problems, with a little bit of work, we can develop a remarkably deep understanding for why certain models behave as they do. Together, my student and I can pick a problem that we think machine learning might be able to offer a solution to. We can explore a variety of methods, experiment with them to see which may offer the best results, and dive into some theory to understand why we are seeing the results that we observe. At the end of our time together I hope that my student will have developed some experience with programming, will have learned some mathematics that they likely won't encounter again until they reach university and will be able to explain the core idea behind several machine learning methods. I would like to conclude our time together by authoring a technical piece of writing (or preparing a technical presentation) which encapsulates the work that we have done together.