|
New horizons in industrial mathematics |
|
|
|
|
Written by Agnar Höskuldsson, DTU
|
During the last decades a new form for mathematics has gained success in analysing industrial data. The basic idea has been to build the solutions to the mathematical models in steps in order to optimise the predictions ability of the final solutions.
At each step the improvement of the solution is judged in two ways, both in the light of the present task (fit, optimisation etc) and also, how precise the solution is for the  given data. The developed methods have had business success both at institutions and industry. The success is partly due to that the new methods secure better predictions than traditional methods and partly that they provide with graphical analysis of data that effectively shows the inherent variation in data. The tools presented here are now working tools within the field of chemometrics.
|
Tutorial