Practical Guide To Machine Learning
H2O has a number of useful white papers, some of which I’ve already blog links too. “Practical Guide To Machine Learning” dropped into my mail box this morning. Although not new, the paper does remind us of the benefit of venturing down a machine learning road if you are willing to put in the time to understand your data, and figure out the algorithms (over 150 classification algorithms are available today) that can help answer the business questions.
IDENTIFY A BUSINESS PROBLEM. Identify opportunities in your business where improved
predictions will have a compelling impact, in the form of increased revenues, reduced costs or some other key business driver. Possible examples include (but are not limited to): detecting and preventing fraud; detecting security risks and threats; measuring credit and default risk; and other high-impact problems. If you can’t find problems like this in your business, you’re not looking hard enough; every business has opportunities to improve.