search

mtl-brainhack-school-2019 / Isabelle-Simard_Multimodal_ML

Project repository
MIT License
1 stars 1 forks source link

Ensemble methods: combining predictors #3

Open atrophiedbrain opened 5 years ago

atrophiedbrain commented 5 years ago

A tutorial here details how ensembles can be created using scikit-learn:

https://towardsdatascience.com/ensemble-learning-using-scikit-learn-85c4531ff86a?gi=f8603ef652d0

After you have a predictor for FNC, power spectra, and neural activity, you'll use an ensemble method to create a predictor which combines 2 or all 3 of these individual predictors.

Unfortunately the tutorial is for classifiers, but it should be relatively straightforward to apply to predictors as well.

Let me know if you have any issues with creating an ensemble for your predictors.

Isabelle-Simard commented 5 years ago

Thank you so much Joshua! I was wondering how I would combine my predictors. I'll have a look at it and let you know if I have issues.

kaufmannSpace commented 5 years ago

Great suggestion @atrophiedbrain, I was about to issue a similar recommendation. You should definitely look into other ML techniques than just regression models and combine them.