Error "OptimizeHyperparameters is not a valid parameter name." while using fitcsvm(X,​Y,'Optimiz​eHyperpara​meters','a​uto').

I have the "Optimization Toolbox" and "Statistics and Machine Learning Toolbox" and MATLAB2016a. Is it a license issue, MATLAB version problem or something else? I could run it without any problem on another PC with MATLAB2017.

Answer 1

[Alan Weiss]

As stated in the Release Notes, Bayesian Optimization was introduced in R2016b.

Alan Weiss

MATLAB mathematical toolbox documentation

[Milad Malekzadeh]

Thanks! I updated my MATLAB to 2017a and the problem is now solved!

Answer 2

[Pravindra Shakunt]

please, can someone provide me latest version of MATLAB for windows and 32-bit operating system. because i don't have license.

[Walter Roberson]

Yes, you can buy a license from Mathworks.

It might no longer be possible to buy a license for R2015b (last version that supported 32 bit windows.) You will probably need to be a license for the current release. If you are a student, Mathworks might not be willing to make R2015b available to you. You might need to purchase a Standard license: with a Standard license, you can definitely download R2015b and use it with your R2018b license. Educational licenses can also (usually) get access to older versions. I have not been able to determine whether Home licenses can download older versions.