I have an idea of how ADABOOST will be used for classification but I want to get the idea of how to re-weight and thus use ADABOOST in case of regression problems.
Here are link to some famous boost of regressor.
- Adaboost.R2: Improving Regressors using Boosting Techniques
- Adaboost.RT: Experiments with AdaBoost.RT, an Improved Boosting Scheme for Regression
Scikit-Learn have many implementations:
As a general principle Adaboost builds and ensemble by sequentially adding members which have been trained on those instances of data which are proving most difficult to correctly classify/predict.
Each new classifier/predictor is given a training set where the difficult examples are increasingly represented, this is achieved either through weighting or resampling.
There should be very little difference in the approach for classification or regression.
AdaBoost is a meta algorithm, so the underlying principle is the same: i would suggest going through this for references.