Venue: ICML 2020 Workshop on Uncertainty & Robustness in Deep Learning (UDL)
Author: Zayd Hammoudeh and Daniel Lowd
This repository contains the source code associated for the UDL'20 paper entitled "Positive-Unlabeled Learning with Arbitrarily Non-Representative Labeled Data"
Official version of the paper published at NeurIPS'20. See the official and latest version of the aPU learner repository. Please open issues in that repository.
To run the program, enter the src directory and call:
python driver.py ConfigFile
where ConfigFile is one of the yaml configuration files in folder src/configs. If CUDA is installed on your system, the program enables CUDA execution automatically.
The first time the program is run, it will download any necessary dataset and create any transfer learning representations automatically. Please note that this process can be time consuming --- in particular for 20 Newsgroups where creating the ELMo-based embeddings can take several hours.
These downloaded files are stored in a folder .data that is in the same directory as driver.py.
Results are printed to the console. The tool also creates a folder named res in the same directory as driver.py where it exports results in CSV (comma separated value) format.
Our implementation was tested in Python 3.6.5. Minimum testing was performed with 3.7.1 but requirements.txt may need to change depending on your local Python configuration. It uses the PyTorch neural network framework, version 1.3.1 and 1.4. For the full requirements, see requirements.txt in the src directory.
We recommend running our program in a virtual environment. Once your virtual environment is created and active, run the following in the src directory:
pip install --user --upgrade pip
pip install -r requirements.txt
MIT.
This repository includes an implementation of PUc [1] that was provided by the tool's author Tomoya Sakai.
[1] Tomoya Sakai and Nobuyuki Shimizu. Covariate shift adaptation on learning from positive and unlabeled data. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 33, pp. 4838-4845, July 2019.