Skip to main content

Introduction To Classification Model Training

In earlier sections for Data Set and Catalog, we saw a few ways of Classification. These are listed as follows:

  1. Classify a Data Set from the Data Set List view.

  2. Give Feedback to predictions of Data Set columns in the Attributes tab of a Data Set.

  3. Give Feedback to Catalog ad-hoc mappings (Mapped Data Sets, Mapped Data Columns).

  4. Give Feedback to Catalog ad-hoc mappings (from Semantic Object Technical View).

    Now let us look at another workflow, where users can manually add new or remove prior mappings at the Concept or Semantic Object level, apart from just giving feedback to existing predictions, which is a more direct way of training the Classification model. This corresponds to the 5th way of Giving feedback:

  5. Give Feedback by adding/removing mappings for a specific Semantic Object or Concept:

    a. Through Train model workflow
    b. Through the Import Concept Mapping feature

And yes, there are even more ways but we’ll take them up, apart from recapping these, in the appropriate sections.