MATLAB: How to count kfoldloss error from ClassificationLinear

bayesoptfitclinearkfoldlosslogistic regressionMATLAB

cv = cvpartition(numel(y_trainUndersampled),'Kfold',5); 
hyperOpt = struct('AcquisitionFunctionName','expected-improvement-plus',...  
            'Optimizer','bayesopt','MaxObjectiveEvaluations', 100,... 
            'CVPartition', cv); 
bestLogsMdl = fitclinear(X_trainUndersampled, y_trainUndersampled,... 
    'Learner', 'logistic',...
    'OptimizeHyperparameters',{'Lambda','Regularization'},...
    'HyperparameterOptimizationOptions',hyperOpt,...
    'ScoreTransform','logit');
Hi, I have used hyperparameter optimization on fitclinear function. The code above produces bestLogsMdl as ClassificationLinear.
I want to use ClassificationLinear to count the kfoldLoss.
However based on the documentation in https://uk.mathworks.com/help/stats/fitclinear.html#bu5mw4p , kfoldLoss is used on ClassificationPartitionedLinear
How to use hyperparameter optimization with fitclinear together on the kfoldLoss? What modifications are needed on the fitclinear so it would produce ClassificationPartitionedLinear?
My ultimate goal is to plot a misclassification rate vs number of learning cycles graph

Best Answer

https://www.mathworks.com/help/stats/fitclinear.html#namevaluepairarguments
You cannot use any cross-validation name-value pair argument along with the 'OptimizeHyperparameters' name-value pair argument. You can modify the cross-validation for 'OptimizeHyperparameters' only by using the 'HyperparameterOptimizationOptions' name-value pair argument.
So you need to get rid of OptimizeHyperParameters and set appropriate HyperparameterOptimizationOptions
Related Question