European Coatings Journal - 15.07.2020 | Kevin Cremanns | ARTIFICIAL INTELLIGENCE

A COMPUTER LEARNS COLOURS

In the development of paint formulations, such as the optimal pigment mixture to produce a colour impression, a classic approach is often performed. This means a previously defined number of experiments is planned and then carried out. The number of experiments depends on the varying factor n [1, 2] and, if applicable, which type of analysis is carried out afterwards. For example, when using a polynomial regression model, the number of experiments N with increasing order m of the model, and a systematic fully factorial design plan result in N=(m+1)n [2]. Should a sensitivity analysis be performed after this, e.g. ...
Want to read more?
Log in or register now!
In order to access this content, please log in or register. Want to test EC 360° first? Start your one month free trial now!
Log in
Free trial access
Buy