notation on linear regression with regularization on Andrew Ng's Machine Learning course

by mLstudent33   Last Updated April 15, 2019 16:19 PM

I was looking here at bias and variance: https://www.coursera.org/learn/machine-learning/lecture/4VDlf/regularization-and-bias-variance

And saw this:

regularization pic

There are two summations with two m's but I cannot figure out what they are. Either the total number of parameters or total number of data points but they should be the same or it makes more sense to use different notation say m and n?

It looks like the left summation is over the data points since x and y are superscripted by i, so in this case m means total number of data points where as the right summation looks like it's over the number of parameters since theta is subscripted by j so in this latter case, it seems like m means the total number of parameters.

I apologize for a rookie question.



Related Questions




GLMs must be 'linear in the parameters'

Updated March 18, 2016 08:08 AM


Is regularization of a linear model really needed?

Updated August 14, 2018 13:19 PM