Page 410 -

guiDe Data Mining in the Real World

“Now suppose I give the names of the three customers any good. We had too many problems with data: wrong, 409
to a salesperson who calls on them, and sure enough, we dirty, and missing. There was no way we could know ahead
have a stream of bad luck and none of them buys. This bad of time that it would happen, but it did.
result doesn’t mean the model is wrong. But what does the “When the time came to present the results to senior
salesperson think? He thinks the model is worthless, and management, what could we do? How could we say we took
he can do better on his own. He tells his manager who tells 6 months of our time and substantial computer resources
her associate, who tells everyone in the Northeast Region, to create a bad model? We had a model, but I just didn’t
and sure enough, the model has a bad reputation all across think it would make accurate predictions. I was a junior
the company. member of the team, and it wasn’t for me to decide. I kept
“Another problem is seasonality. Say all your training my mouth shut, but I never felt good about it. Fortunately,
data are from the summer. Will your model be valid for the the project was cancelled later for other reasons.
winter? Maybe, but maybe not. You might even know that “However, I’m only talking about my bad experiences.
it won’t be valid for predicting winter sales, but if you don’t Some of my projects have been excellent. On many, we
have winter data, what do you do? found interesting and important patterns and information,
“When you start a data mining project, you never and a few times I’ve created very accurate predictive mod-
know how it will turn out. I worked on one project for 6 els. It’s not easy, though, and you have to be very careful.
months, and when we finished, I didn’t think our model was Also, lucky!”

DisCussion Questions

1. Summarize the concerns expressed by this data analyst. was ineffective, maybe even wrong, what would you do?
2. Do you think the concerns raised here are sufficient to If your boss disagrees with your beliefs, would you go
avoid data mining projects altogether? higher in the organization? What are the risks of doing
3. If you were a junior member of a data mining team and so? What else might you do?
you thought that the model that had been developed