A Practical Guide to Data Mining for Business and Industry

13 Overviews

Overviews

1. 13.1 To Make Use of Official Statistics
2. 13.2 How to Use Simple Maths to Make an Impression
   1. 13.2.1 Approximations
   2. 13.2.2 Absolute and Relative Values
   3. 13.2.3 % Change
   4. 13.2.4 Values in Context
   5. 13.2.5 Confidence Intervals
   6. 13.2.6 Rounding
   7. 13.2.7 Tables
   8. 13.2.8 Figures
3. 13.3 Differences between Statistical Analysis and Data Mining
   1. 13.3.1 Assumptions
   2. 13.3.2 Values Missing Because ‘Nothing Happened’
   3. 13.3.3 Sample Sizes
   4. 13.3.4 Goodness-of-Fit Tests
   5. 13.3.5 Model Complexity
4. 13.4 How to Use Data Mining in Different Industries
5. 13.5 Future Views

13.1 To Make Use of Official Statistics

Official statistics are gathered by National Statistics Institutes and are a free resource for businesses and data miners. The quality of official statistics is assured by regulated checking methods, and historic series of matching data are available. Compatibility across countries is also generally good. Official statistics can inform a business about regional hotspots of potential customers, opportunities for specific types of products and services and neighbourhood networks. For example, official statistics are available for population size stratified by age, gender and small geographic areas.

The wealth of information available is remarkable. It is possible to find data on health, employment and education as well as composite data such as distance travelled to work and population flows in and out of areas. If continuous variables are transformed by categorising ...