Have you used sklearn's Voting Classifier? | Coursera Community
Coursera Header

Have you used sklearn's Voting Classifier?

  • 11 December 2018
  • 1 reply

Userlevel 5
Badge +5
  • Participant
  • 59 replies
sklearn has a Voting Classifier which allows you to combine conceptually different ML models (see more here)

Has anyone experimented with this before? Describe the problem you applied it to and how well did it work?

1 reply

Userlevel 3
Badge +5
Hello @Liz

I have not used voting classifier . But it is similar to ensemble method . I came to know about these in deep Learning specialization from deeplearning.ai from Andrew ng . He said about this in 2nd course : Improving neural network. He said these methods are good for kaggle contests or other datascience contest . But if you want to deploy it as working model it is not advisable . Both size of the model and running time of the model increases. Instead you can focus on improving a single model by Fine tuning the parameters instead of training different models. This is my opinion. You may try it if dataset is too small to avoid overfitting by training a single model.