In Linear Regression Model Part 1, Dr. Ng uses a superscript notation for the ith training example.
I found this to be a little surprising, since in lots of math formulae and machine learning books, I usually see the subscript used.
Is this parenthesized superscript style common? Perhaps I am out of date with modern conventions.
In ML, everyone seems to invent their own conventions. There is little standardization.
Hmmm… I agree with your first statement but perhaps not with the second: mathematical notation tends to be quite standardized, no?
Math has very stable notations and conventions. ML not that much, even in the algebra. In fact, in many lessons, Prof. Ng mentions and clarifies how the notation among different people within the ML domain varies. He tries to be very consistent across the board and again, he usually explains clearly his conventions and even presents other people’s conventions.
Fair enough, thanks for the thoughtful replies!