Skip to content

OvaModelParameters is not strongly-typed #2467

Open
@wschin

Description

@wschin

In the constructor of 'OvaModelParameters`,

        using TScalarPredictor = IPredictorProducing<float>;
        ...
        internal static OvaModelParameters Create(IHost host,  OutputFormula outputFormula, TScalarPredictor[] predictors)
        {
        ...
        }

we remove the type information of predictors by forcing them to be IPredictorProducing<float>. To make all multi-class classifiers' outputting strongly-typed models. We need to make OvaModelParameters strongly-typed.

Activity

self-assigned this
on Feb 7, 2019
added
APIIssues pertaining the friendly API
on Feb 11, 2019
added
P0Priority of the issue for triage purpose: IMPORTANT, needs to be fixed right away.
on May 21, 2019
wschin

wschin commented on May 21, 2019

@wschin
MemberAuthor

Without strongly-typed OVA, user can't access the trained models because variable can be accessed when type is known.

assigned and unassigned on May 22, 2019
added
P1Priority of the issue for triage purpose: Needs to be fixed soon.
and removed
P0Priority of the issue for triage purpose: IMPORTANT, needs to be fixed right away.
on Jan 17, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

APIIssues pertaining the friendly APIP1Priority of the issue for triage purpose: Needs to be fixed soon.

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

    Participants

    @shauheen@wschin@najeeb-kazmi@ganik@michaelgsharp

    Issue actions

      OvaModelParameters is not strongly-typed · Issue #2467 · dotnet/machinelearning