[jira] [Created] (IGNITE-11192) [ML] Use nd4j for matrix inversions and determinants

[Anton Vinogradov (Jira)]

Alexey Platonov created IGNITE-11192:
----------------------------------------

             Summary: [ML] Use nd4j for matrix inversions and determinants
                 Key: IGNITE-11192
                 URL: https://issues.apache.org/jira/browse/IGNITE-11192
             Project: Ignite
          Issue Type: Improvement
          Components: ml
            Reporter: Alexey Platonov
            Assignee: Alexey Platonov


From optimization point of view we should use matrix inversions and determinant computations of dl4j instead of own realization.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

[Alexey Zinoviev]

Could you please edit description and write in what tasks/algorithms we are
using matrix inversion /determinant computations (post link to our own
realization and its usages)

пн, 4 февр. 2019 г. в 16:23, Alexey Platonov (JIRA) <[hidden email]>:

> Alexey Platonov created IGNITE-11192:
> ----------------------------------------
>
>              Summary: [ML] Use nd4j for matrix inversions and determinants
>                  Key: IGNITE-11192
>                  URL: https://issues.apache.org/jira/browse/IGNITE-11192
>              Project: Ignite
>           Issue Type: Improvement
>           Components: ml
>             Reporter: Alexey Platonov
>             Assignee: Alexey Platonov
>
>
> From optimization point of view we should use matrix inversions and
> determinant computations of dl4j instead of own realization.
>
>
>
> --
> This message was sent by Atlassian JIRA
> (v7.6.3#76005)
>