Min objective and function evaluations

3 次查看(过去 30 天)
As I was learning to optimize regression tree, I'm struggling to understand some of the codes and graphs generated in the matlab example ' Optimize Regression Tree'
load carsmall
X = [Weight,Horsepower];
Y = MPG;
rng default
Mdl = fitrtree(X,Y,'OptimizeHyperparameters','auto',...
'HyperparameterOptimizationOptions',struct('AcquisitionFunctionName',...
'expected-improvement-plus'))
As you can see from the above code, they set the 'OptimizeHyperparameters' to 'auto', they struct 'AcquisitionFunctionName' to 'expected-improvement-plus', they also put 'HyperparameterOptimizationOptions' in the bracket.
My first question is that i'm not familiar with all the parameters I could put here, is there a list of those parameters out there for me to familiarize with all the properties I could put in the bracket?
Once you type the above code, the outputs are two graphs shown below.
OptimizeARegressionTreeExample_01.png
OptimizeARegressionTreeExample_02.png
My second question is that in the first graph, what does 'Min objective' mean? What does 'Number of function Evaluations' mean?

回答(1 个)

Don Mathis
Don Mathis 2019-1-16
Question 2: As mentioned in the link for Question 1, it's using the 'bayesopt' function. Start here: https://www.mathworks.com/help/stats/bayesopt.html

类别

Help CenterFile Exchange 中查找有关 Model Building and Assessment 的更多信息

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by