CODEWITHSUNDEEP
machine-learningdeep-learningdatasetdata-science
hadeer adel
hadeer adel

Reputation: 11

combining Machine learning models

Is there is a way to combine multiple ML models where each model use a dataset with different features but same outcome?

I find a way to combine models which is ensemble learning but its only applicable for datasets with the same features

Upvotes: 1

Views: 798

Answers (1)

Julia Meshcheryakova
Julia Meshcheryakova

Reputation: 3262

It is possible if you have a way to match the entity fields in different datasets (e.g. same id).

Then you can train your models independently and ensemble them in any fashion (bagging, boosting, stacking)

The easiest to try would be a linear combination. E.g.:

  • 0.5 * y_pred_1 + 0.5 * y_pred_2
  • 0.8 * y_pred_1 + 0.2 * y_pred_2 enter image description here

Also refer to: similar question 1 similar question 2 similar question 3

Upvotes: 2

Related Questions