**Optimization, adaptation and regression
analysis**

*Topics*

1. | Data structures. Knowledge databases. | |

2. | The H-principle of modelling data. | |

3. | Knowledge and data. Measures of prediction. | |

4. | Optimization procedures. | |

5. | Adaptation procedures. | |

6. | Target modelling. | |

7. | Weighing procedures. | |

8. | Key numbers. | |

9. | Data mining. | |

10. | Goal modelling/optimization. | |

11. | 'Best practice' procedures. | |

12. | Knowledge management. | |

13. | Causal interpretation of data structures | |

14. | Subset selection of data | |

15. | Performance measures. | |

16. | Dispersion in data (Light scattering) | |

17. | Knowledge matching. Degree of adaptation. The 80-20 rule. | |

18. | Score values. Score cards. | |

19. | Summary results from analysis. | |

20. | Case study | |

21. | Guidelines for presentation of results |

In industry it is often needed to look at the modelling task in an untraditional way. In industry there is trend of evaluating activities in terms of money involved. Industry is looking for methods that seek to include values in terms of money in the models. The experimenter may not be able to make precise what is the value of one unit of a variable. But business people are getting good training in assessing money values to more and more activities in the company. Here are introduced a collection of methods that can be used to combine optimization procedures with regression type of analysis. Another big problem in industry is the large redundancy in data. Under the name of data mining a collection of methods have been developed that test out different approaches to find out if there is 'something' in data and in case there is, then what is it. A more efficient procedure is to use a knowledge base as a guidance tool, and let the H-methods find out what gives predictive performance.

See a short review of the ideas of mixing regression and optimization.

**
**