Data mining is a relatively new branch of computer science that involves looking for meaningful patterns in large data sets. It can involve a combination of traditional and new data analysis techniques, including standard statistical tools and machine learning algorithms.
The particular data mining tools used in given situation depend upon factors such as:
- What types of data are available
- How much (i.e., how many individual records) of data is available
- The marketing-related (or other type of) goals of the project
Data mining excels where traditional statistics leaves off. It is particularly useful and appropriate when the following conditions are present:
- The project involves the analysis of large data sets that potentially contain a lot of noise (or erroneous data)
- The project designers or business managers are unsure as to which variables might be predictive of a particular outcome
- There is a desire to group like records together based upon some attribute or variable
The result of a data mining exercise can include the classification of certain subsets of the data into new, as-yet-unforeseen categories. It can also result in the determination of which variables or combinations of variables matter in terms of being predictors for a given outcome ? and how much.
Contact MindEcology today to get started.