what does the plot obtained from fitcsvm mean?

Hi, I am using fitcsvm for classifying my data, and I am using the optimization operation to auto-select the kernel, and scale. Using that, I obtain a plot of Min objective vs the number of function evaluations, I did not understand the plot, can someone please help me in that. Thanx

Answer 1

[Shrestha Kumar]

Hi,

In the plot, the number of function evaluations is the iteration number of the objective function(i.e first iteration,second iteration and so on). The minObjective is the minimum value of the objective function reached in that iteration.

Here the objective function is the logarithm of one plus the cross-validation loss.

If you want to read about the optimization technique used here (Bayesian optimization) then you can refer to this link -

https://in.mathworks.com/help/stats/bayesianoptimization.html

[Foram Sanghavi]

Thank you so much, now it makes sense.