Package description
IDEAL is a Python package implementing an active-learning method for regression based on inverse-distance weighting functions for selecting the feature vectors to query. The algorithm is independent of the type of predictor used and can handle known and unknown constraints. The active learning function accepts predictors in scikit-learn format and can be used also for classification problems.
The algorithm is based on the following paper:
[1] A. Bemporad, "Active Learning for Regression by Inverse Distance Weighting," arXiv eprint 2204.07177, 2022.
This software is distributed without any warranty. Please cite the above paper if you use this software.
Download
Latest version:
IDEAL v3.0 - December 11, 2022
(additional support for improved representativeness-diversity maximization)
Previous versions:
IDEAL v2.0 - October 27, 2022
(additional support for: random sampling, greedy sampling based on distances between feature vectors, greedy sampling based on feature vectors and predicted targets, query-by-committee for regression)
IDEAL v1.0 - April 14, 2022
(first public release)
Last update: December 11, 2022