You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have a KRR model, and I checked that it is differentiable computing numeric derivatives.
Although the calculation is fast, I would like to know whether there is a way to get analytic or automatic derivatives of the function with respect to input features, besides me doing the math for said derivatives.
To give more information, each input feature is an inverse euclidean distance $\frac{1}{r_{ij}} = \frac{1}{\lvert r_i - r_j \rvert}$ between two points, so I would actually need the derivative with respect to cartesian displacements $\left( \frac{\partial f}{\partial x_i}, \frac{\partial f}{\partial y_i}, \frac{\partial f}{\partial z_i} \right)$.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
Hello,
I have a KRR model, and I checked that it is differentiable computing numeric derivatives.
Although the calculation is fast, I would like to know whether there is a way to get analytic or automatic derivatives of the function with respect to input features, besides me doing the math for said derivatives.
To give more information, each input feature is an inverse euclidean distance$\frac{1}{r_{ij}} = \frac{1}{\lvert r_i - r_j \rvert}$ between two points, so I would actually need the derivative with respect to cartesian displacements $\left( \frac{\partial f}{\partial x_i}, \frac{\partial f}{\partial y_i}, \frac{\partial f}{\partial z_i} \right)$ .
Beta Was this translation helpful? Give feedback.
All reactions