Machine Learning for Finding Patterns in Data
Computers are truly awesome at spotting patterns in data. You could say it’s the app they were born to serve. Computers can process millions of rows of data, while the processing speed of a human is limited to their ability to scan and interpret data through their eyes. If you’ve ever tried to make sense of data presented in rows and columns on a spreadsheet, you will appreciate how difficult it is for humans to spot patterns, while such a task is quick and painless for a computer. Modern enterprise system designs seek to blend these powerful pattern recognition skills found in computers with the more obscure interpretive competencies of humans to make better decisions, and more often.
An example of this was found in the slot machine gambling industry. A large casino in Las Vegas monitored the movement of its customers around its main gambling hall, and a machine learning algorithm spotted changes in the behaviors of older customers. Their computer system highlighted that, in the late afternoon, gamblers over 55 would huddle around a particular set of slot machines, while other slot machines went unused. After investigation, the owners realized the reason for this "pattern" was the desire of older folks to want to have the sun on their backs in the late afternoon while they were using the slot machines. This resulted in a re-design of the hall that substantively increased revenues by relocating slot machines to make more of them available for use.