Difference between revisions of "MultipleRunFItter"
| Line 1: | Line 1: | ||
Let's look at the [[Media:Lab_R.pdf|homework]] due last night. | Let's look at the [[Media:Lab_R.pdf|homework]] due last night. | ||
| − | Today we are going to use NEURON for System ID, specifically to fit a function to data. The function represents a synaptic conductance. | + | Today we are going to use NEURON for System ID, specifically to fit a function to data. The function represents a synaptic conductance. |
| + | |||
| + | Today's tutorial will help prepare us for a lecture on Optimization. | ||
Fitting functions is simpler than fitting models. Consider that the values function we are fitting today depend only on time. The values of a model simulation depend upon the history of the model. There is another tutorial for fitting models which we may get to. | Fitting functions is simpler than fitting models. Consider that the values function we are fitting today depend only on time. The values of a model simulation depend upon the history of the model. There is another tutorial for fitting models which we may get to. | ||
Latest revision as of 16:53, 7 April 2009
Let's look at the homework due last night.
Today we are going to use NEURON for System ID, specifically to fit a function to data. The function represents a synaptic conductance.
Today's tutorial will help prepare us for a lecture on Optimization.
Fitting functions is simpler than fitting models. Consider that the values function we are fitting today depend only on time. The values of a model simulation depend upon the history of the model. There is another tutorial for fitting models which we may get to.
First, download the data to be used to fit the model.
Now go here and do the first tutorial.
Homework S.1: Go through today's tutorial. Perform the optimization and paste the plot of the "fit" function overlaying the "data" function into PowerPoint.
Homework S.2: Answer the questions under the heading "Extra credit" problem at the end of the tutorial.
