Submitted by Gary Short
The clue to the issue with machine learning is in the name; machine learning, i.e. the machine learns from past experience. But what happens if past experience is not a good predictor of what will happen in the future?
I'll demonstrate this point with a simple horse race. Just because Horse A won a race last week, doesn't mean it will win this week. So what can we do? During this talk I'll walk you through a solution to this problem, focusing on the concrete example of a horse race. Not only that, but we'll take what we learn and apply it to a race being run at the same time as my talk, and we'll (hopefully) predict the winner. :-)
Gary Short is a Data Solution Architect for Microsoft. He specialises in machine learning and “big data” on the Azure Platform, but has an interest in data science, in all forms, especially computational linguistics and social network analysis.