About coefficient in results of summary command, when using linear regression algorithm in MLTK.

[yutaka1005]

When linear regression algorithm is used in MLTK, the field called coefficient is returned by the summary command.

Is this value a Partial regression coefficient?
Or Standard Partial regression coefficient?

If the former is correct, I think that because the dimensions of each variable are different (price, distance, percentage, etc.), the magnitude of those values is unreliable and I can't know which variable is important.

If anyone knows about it, please tell me.

[aoliner_splunk · Accepted solution]

We are returning the coefficients provided by scikit-learn, specifically the .coef_ attribute, as documented here:
http://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LinearRegression.html

You can inspect the code directly in the MLTK directory under ml-spl/bin/algos/LinearRegression.py

[yutaka1005]

Thank you for answer.

Since the command summary seems to get the attribute of coef as it is from sklearn, the field coefficient seems to be a partial regression coefficient not standard partial regression coefficient as far as I checked ml-spl/bin/algos/LinearRegression.py and sklearn.