Machine Leaning with Tinkerpop?

Are there any groups trying to implement machine learning on top of tinkerpop?

Here are some methods that would support graph based machine learning:

a. Project a subgraph to a sparse matrix.
b. Project a subgraph to a sparse graph Laplacian matrix.
c. Sparse matrix eigenvalues and eigenvectors using bsp.
d. Sparse matrix factorization using bsp.
 i. QR factorization.
 ii. LU factorization.
 iii. Choleski factorization.
 iv. Non negative factorization.
e. Convex optimization
 i. ADMM solver using bsp.
f. Probabilistic Soft Logic

I was wondering if any of these methods had been implement on top of the compute engines with bsp.

Sincerely,
David G. Boney
david.g.boney@gmail.com

Re: Machine Leaning with Tinkerpop?

[Stephen Mallette <sp...@gmail.com>]

I'll just answer your basic question at this point and say that I'm not
aware of anyone working on those kinds of things with TinkerPop in mind.

On Sat, Mar 31, 2018 at 10:06 PM, david.g.boney@gmail.com <
david.g.boney@gmail.com> wrote:

> Are there any groups trying to implement machine learning on top of
> tinkerpop?
>
> Here are some methods that would support graph based machine learning:
>
> a. Project a subgraph to a sparse matrix.
> b. Project a subgraph to a sparse graph Laplacian matrix.
> c. Sparse matrix eigenvalues and eigenvectors using bsp.
> d. Sparse matrix factorization using bsp.
>  i. QR factorization.
>  ii. LU factorization.
>  iii. Choleski factorization.
>  iv. Non negative factorization.
> e. Convex optimization
>  i. ADMM solver using bsp.
> f. Probabilistic Soft Logic
>
> I was wondering if any of these methods had been implement on top of the
> compute engines with bsp.
>
> Sincerely,
> David G. Boney
> david.g.boney@gmail.com
>