DESIGN OF EXPERIMENTS for nonlinear models and neural networks.
OPTIMIZED DESIGN OF EXPERIMENT
MINIMIZING THE NUMBER OF TRIALS
BETTER ACCURACY FOR CONFIDENCE INTERVALS

Les
designs of experiments are magnificent tools to:
- Better understand the studied phenomenon,
- Optimally organize the trials,
- Obtain as much information as possible with less trials,
- Minimize the number of trials, if it is a priority,
- Obtain the most accurate confidence intervals for the studied model,
- Calculate the most accurate predictions.
The history of design of experiments began with Fisher in 1929.
NETRAL is the software editor in the world to propose a software, NEURO PEX, dedicated to design of experiment for non-linear models and neural networks. Of course, NEURO PEX can deal with all optimal designs for linear models as well.
NETRAL is also the first society to offer consultancy in design of experiments for phenomena that are described by non-linear models or by neural networks.

Discover the qualities of
NEURO PEX.

Please
contact us if you are interested in our offer of consultancy in design of experiment
for nonlinear models and neural networks.

Discover the
Box-Lucas example (knowledge-based model) handdled as a 51-screenshot tutorial.

Discover the
Porphyridium example (knowledge-based model).

Discover an example of
design of experiments for neural networks.

The
Designs of computer experiments Day took place the 24th of March 2005 in the Institut Henri Poincaré.

Visit the website of the
Expérimentique association from Orléans, where French-speaking design of experiments
specialists meet each other.
© Netral - June 2007