Trying to find ideal parameters for a function.

So I have many-channel electrical recordings. These recordings have events in them that I'm trying to automate the detection of. I found a function to detect the events, but it requires passing in some parameters. I've fooled with them a little bit to try and get the function to output results that I agree with for a few recordings, but ultimately I have no idea if what the function is calling an "event" is something that I would myself call an event.

What I'm looking for is a less subjective way of setting the parameters that I pass into this event detection function. I'm imagining something like:

1. Go through a few recordings and do event detection myself (count number of events/recording)

2. Run the detection function many many different times on those same few recordings with different parameters each time in some kind of systematic way. So it might look like EventDetection(data, i, j, k) and then I change i j and k every time.

3. Subtract my numbers from the function's numbers each iteration and find the iteration where that difference is most reliably close to zero.

Is there a better way of doing this than just looping through a bunch of times?