Gordon MacMillan
Galvanize Data Science Immersive - Case Study - March 2017
Say you run a website or app where you host users events which allows the user to see who is coming, share updates, sell tickets and keep track of banners, descriptions, social media and all manner of event data. This website is doing great except a number of visitors to user sites are complaining of fraudulent events.
These fraudulent events may be b2b sales conferences, DJ shows where exclusive celebrities are expected to perform or even "Get Rich Quicker" book sales schemes. You are interested in having a model that predicts the probability that a new event could be fraud based on a number of predictor features of that event. I'm the data scientist that gets to make this model.