Data Mining For Dummies

Appendix A

Glossary

analysis: Thoughtful investigation of real-world systems.

analytics: Analysis that involves math. (This term is used very differently by different people, and may refer to anything from simple historical data summaries to highly complex predictive models. Always ask questions!)

association rules: Tools for identifying combinations of items often found together. The most common use of association rules is for market basket analysis.

assumption: Something presumed to be true. Assumptions are the basis of all statistical analysis. (It is important that the analyst choose methods based only on assumptions that are reasonable for the application.)

average: Any measure that describes the middle (more formally, “central tendency” or “location”) of a distribution. In analytics, the term “average” usually refers to the mean, but may refer to median or mode.

Bayesian network: A type of neural network. The Bayesian network is based on the fundamentals of probability theory. (See also neural network.)

binary: Having exactly two alternative states.

binning: Organizing data into groups. This may be done for ease of analysis, or to protect privacy.

causation: The act of producing an effect or making something happen. The phrase “correlation does not imply causation” means that the fact that two things are observed to happen together is not enough to prove that one caused the other.

Chi-square: A test statistic, probably the most widely used of all statistical hypothesis ...