Dramatic increases in computing power and storage allow more and more data to be stored. These collected data offer immense opportunity for analysis, but can often be overwhelming without proper Data Mining techniques. We at Mason Statistical Consulting group are eager to help our clients make full use of their data by applying Data Mining techniques to identify consistent patterns and/or systematic relationships between variables, and then to validate the findings by applying the detected patterns to new subsets of data.
Using SAS© Enterprise Miner, there are both predictive and classification methods. In case of a binomial outcome, e.g., an individual “will buy” or “will not buy”, Enterprise Miner predicts the probability of “will buy” from other characteristics of the individual. For a continuous outcome, Enterprise Miner predicts the value of the outcome given other characteristics of the individual. With categorical data, a cluster analysis may be developed in order to assign an individual to a particular cluster. Using Text Miner, documents may be sorted by “type” of document, e.g., for legal documents, sorted by areas of law.
