5. Evaluations
The fitting depends on the number of fitted parameters. In general, the more fitted parameters the model assumes, the more sensitive outcome of the fitting is on their initial value and consequently their initial accuracy needs to be higher.
The decreasing lambda value indicates a successful convergence of the fitting procedure. The sum of squares measures the difference between the measured and the fitted values of the dependent variable.
It is important to choose the model based on experimental information, because a given set of data can generally be fitted to multiple models.
Starting Evaluations
To start the evaluation press the green Start button:
Your job request will be sent to the server, queued and processed. The Start button is disabled in the meantime.
If the server is busy, Waiting ... and the position
of your request in queue will be displayed in the Evaluated Parameters field:
Requests are processed as they reach position 1.
When the evaluation starts, evaluated parameters and curve fitting are shown for each iteration approximately every second during the
evaluation.
Stopping/Cancelling Evaluations
You may stop the evaluation at any time by pressing the red Stop button. This also applies for evaluation requests waiting in queue. The Stop button is enabled only upon pressing the Start button.
Evaluation Output
When an evaluation finishes, either by converging or by reaching the iteration limit, the final evaluated parameters will be shown.
These represent fitted kinetic parameters that best describe the concordance between theoretical curves of the proposed kinetic model
and the experimental kinetic data.
If the convergence is reached, the proposed model is considered to be plausible and confirmed, otherwise you are encouraged
to repeat the calculation with a different set of initial parameters and species concentrations.