Just for clarification (because we didn't do a example in class):

The hypothesis function multiples each row of our features matrix by the theta vector.

After that process we subtract our actual value y_i from the vector that returns from the hypothesis (this is a 1x1 single value).  Y is a vector, so y_i should be a single value as y in this case is a 415x1 vector.  Correct?

We multiply it by the currect x_j value and repeat this process for all x_ j...

Also dumb question, but do we also need to scale our actual values y ?