Data Mining for the Midsize Business

In the large enterprise we constructed a number of datamining efforts that looked to classify solutions based on measurements we had gathered.   One popular technique, called Regression Tree Classification, made it easy to create and adapt decision trees based on small amounts of data.

Decision trees are very easy for management to understand.   I would often be approached with a data set that would be much smaller than the 'Big Data' being gathered today, and quickly construct a model that could be tested against future data.

We used this method against many applications.   Many other such modeling methods exist and are available in systems like R.  I am now in the midst of addressing such a datamining problem for a midsize startup.

I see in a recent IBM Midsize Insider article about how to think about setting up and utilizing a data mining program.    The piece covers the issues very broadly.  I would suggest also to include an expert consultant in data mining, at least for the first few examples.  Also, make sure to start simple and to start with problems that are likely to provide quick results.

Further, include business decision makers in the selection of the data, ensuring that it is correct, and in the use of the decision recommendations in the real world.

The article linked to above also provides a good list of application ideas that are good to examine for initial applications.   If you gather information about how you are doing your business, and seek methods that will make it operate better, take a look at data mining methods.  I often cover data mining in this blog.

This post was written as part of the IBM for Midsize Business program, which provides midsize businesses with the tools, expertise and solutions they need to become engines of a smarter planet. I’ve been compensated to contribute to this program, but the opinions expressed in this post are my own and don't necessarily represent IBM's positions, strategies or opinions.