Gradient descent: Difference between revisions
Content added Content deleted
Line 213: | Line 213: | ||
Note the different implementation of <code>grad</code>. I believe that the vector should be reset and only the partial derivative in a particular dimension is to be used. For this reason, I've _yet another_ result! |
Note the different implementation of <code>grad</code>. I believe that the vector should be reset and only the partial derivative in a particular dimension is to be used. For this reason, I've _yet another_ result! |
||
I could have used ∇ and Δ in the variable names, but it looked too confusing, so I've gone with |
I could have used ∇ and Δ in the variable names, but it looked too confusing, so I've gone with <var>grad-</var> and <var>del-</var> |
||
<lang racket>#lang racket |
<lang racket>#lang racket |