The virtues of machine learning: When statistical analysis isn't enough

By sheer virtue of the number of our waking hours we spend working, it wouldn't be unreasonable to say that many of these decisions are work-related. Some of them are simple, such as "should I respond to this email now, or in a few hours?" Others are much more open-ended, and require many small decisions before an adequate big decision can be arrived at: "How can we improve the likelihood that our clients will buy into our ancillary services?" The simpler work decisions are almost automatic: "I have to go to a meeting now, so I'll respond to this email later." But this isn't how bigger decisions, such as the quandary mentioned above, are made.