-
Try with the lion optimizer. I've found these much better for piecewise polynomial networks, then again, maybe it won't be a huge benefit in convolutional models...
-
There are sooo much _polynomial_ KANs, beside this there are Chebyshev, Legendre, etc...
After implementing several of these KANs, I wonder
- is this really meaningful?
- are these types of _poly…
-
Hi! I read your paper "Deep Polynomial Neural Networks", and I want to know the polynomial networks architecture you implement on image classification. It seems that there is no the folder about imag…
-
Reading through https://arxiv.org/abs/2311.08610 and I wanted to highlight some of the polynomial approximation techniques that we could incorporate into HEIR when the time is right:
> 2.3 Polynomi…
j2kun updated
10 months ago
-
### Evaluation
- [ ] [Confusion matrices](https://towardsdatascience.com/understanding-confusion-matrix-a9ad42dcfd62)
- [Source 2](https://en.wikipedia.org/wiki/Confusion_matrix)
- [ ] [Precision…
-
Hello,
I am not sure whenever this is possible or not?, but I am interested in creating polynomials with parameters.
Say that I have a polynomial with variables
```julia
@polyvar x y
```
and …
-
I have another question on the **strict requirement** of a **diagonal** `A` arising from some mathematical relationship with glorot/xavier initialization for `A`.
![image](https://github.com/state-…
-
### ML-Crate Repository (Proposing new issue)
:red_circle: **Project Title** : Telecom Dataset Analysis
:red_circle: **Aim** : Create a bunch of ML models to analyze the telecom dataset.
:red_circl…
-
**What would you like to submit?** (put an 'x' inside the bracket that applies)
- [x] question
- [ ] bug report
- [ ] feature request
**Issue description**
Hi,
Are there any plans …
-
When approximate the initial filters kernel gθ(Λ) = diag(θ) with a polynomial
![image](https://user-images.githubusercontent.com/40826585/44521567-1512dc00-a706-11e8-8219-2b85f0f71002.png)
why we c…