I would like to find the solution of a system of equations of the form:

$A = w_1 F(x_k) + w_2 F(x_l) + w_3 F(x_m) +...$,

where the unknowns are the $w_i$ and the function $F$, while my data are the $A$ and $x_i$s. I have a system of those equations for different $A$s and different points $x_i$ where the function $F$ is evaluated.

The plan is to expand the function $F$ on some basis.

I am wondering whether there is some numerical method specifically designed for problems with this kind of structure.

One possible solution would involve iterations with an alternating least-squares between optimizing the $w_i$s while holding the $F$ fixed and then optimising the function approximation for fixed $w_i$s. Yet I am not sure whether this approach would converge.

Is anyone of you aware of something along these lines?


1 Answer 1


The equation described here looks familiar to many machine learning approaches.

It appears to fit the paradigm of a regression equation where you have designed/engineered "features" by applying a function, $F$, to your raw data, $x_i$.

The $w_i$'s would be the weights that can be learned by least squares or other methods.

If you re-use the function $F$, this problem could be structured as a single hidden-layer neural network. In this case $F$ would typically be a function of the group called activation functions.

It would be ideal to explore the addition of an intercept/bias term to the equation for A.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.