The repository for implementation of BALPI framework
This work proposes a new framework for phase identification tasks with BAL.
We formulate the phase identification task as two different styles: classification and level-set estimation by considering the target label as classes or utilizing scores to indicate the phases, respectively.
For the predictive models, we consider Gaussian Process Classifier (GPC) and Gaussian Process Regressor (GPR) respectively. We also consider different utility functions for the classification formulation and use Straddle (termed as UCB in code) to guide sample selection for the level-set estimation formulation
The code can be found in the folder code and we prepared two example data in the folder data for the regression data and the classification data,respectively.
Please check the main.py file for an examplar runable code.
For the required packages, please see the requirement.txt for the prerequisite packages or simply use
pip install requirements.txt