|
|
Knowledge Generation
Tasks
Task Number tables: The common way of gathering data. Oftentimes a "Data cemetery" is created without further generation of information or even knowledge. Solution To create conclusions and knowledge out of tabelled numbers using statistical multivariate data analysis. Or to deduce rules from facts and figures with the assistance of inductive methods (ID3-Algorithm). Or to build complex mathematical models on the basis of neural networks.
Task Research / Development: "Keep-everything-constant-and-vary-one-parameter" is a very common development strategy. Imagine a blindfolded person who tries to hit the bull's eye ... Solution The parameters are varied simultaneously according to a statisically balanced plan. By doing so, you can generate a maximum information yield with a minimum expenditure. The goal is being achieved with Experimental Design and computer aided data analysis.
Task Production / Optimisation / Quality control: Very often an engineer would like to improve product yield and simultaneously reduce by-products or costs. Or somenone finds variations in quality but does not discover its cause. Or simply, someone wants to determine the influence of a specific parameter. Solution Stepwise optimisation following the EVOP-method (Evolutionary Optimisation). The most important paramaters can be identified and separated from the less important ones. Additionally, you may find synergistic effects.
|