Error occurred in fitlm function.

Error using classreg.regr.TermsRegression.createFormula

The terms matrix must have one column for each variable in the dataset or table. You may not

specify the response as a separate vector with a dataset or table.

Error in LinearModel.createFormula (line 1076)

formula = classreg.regr.TermsRegression.createFormula(supplied,modelDef, ...

Error in LinearModel.fit (line 1058)

model.Formula = LinearModel.createFormula(supplied,modelDef,X, ...

Error in fitlm (line 122)

model = LinearModel.fit(X,varargin{:});

Error in reg_VI (line 45)

reg = fitlm(X_train,Y_train);

[Walter Roberson]

I speculate that your Y_train might not be a column vector

[Walter Roberson]

It is possible that Y_train has multiple columns.

[the cyclist]

@Devendra, can you upload the X and Y data? (You can use the paper clip icon in the INSERT section of the toolbar.) This would make it vastly easier to help you debug this.

Answer 1

[the cyclist]

@Walter Roberson surmised correctly, as usual. The fitlm function cannot be used when you have multiple dependent variables.

You may be able to use mvregress. (It won't be a machine learning algorithm, though.)