Positive and Unlabeled Materials Machine Learning (pumml) is a code that uses semi-supervised positive and unlabeled (PU) machine learning to classify materials when data is incomplete and only examples of 'positive' materials are available.
Positive and Unlabeled Materials Machine Learning (pumml) is a code that uses semi-supervised positive and unlabeled (PU) machine learning to classify materials when data is incomplete and only examples of "positive" materials are available. As an example, pumml was used to predict the "synthesizability" of bulk and 2D materials from "positive" examples of synthesized materials.
How to cite pumml
If you use pumml in your research, please cite the following work:
Nathan C. Frey, Jin Wang, Gabriel Iván Vega Bellido, Babak Anasori, Yury Gogotsi, and Vivek B. Shenoy. Prediction of Synthesis of 2D Metal Carbides and Nitrides (MXenes) and Their Precursors with Positive and Unlabeled Machine Learning. ACS Nano 2019 13 (3), 3031-3041.
Please also consider citing the original works that establish the underlying methodology of pumml:
Elkan, Charles, and Keith Noto. Learning classifiers from only positive and unlabeled data. Proceeding of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining. ACM, 2008.
Mordelet, F.; Vert, J.-P. A Bagging SVM to Learn from Positive and Unlabeled Examples. Pattern Recognit. Lett. 2014, 37, 201−209.
The easiest way to get started with pumml is to create a virtual environment with python3.6 and then
pip install pumml
You can also create a virtual environment, clone this repo and do
python setup.py install in the root directory.
An example Jupyter notebook called
example_nb.ipynb shows the basic functionality of the package.
More information about using PU learning for materials synthesis prediction can be found in our publication:
DOI: 10.1021/acsnano.8b08014 https://pubs.acs.org/doi/abs/10.1021/acsnano.8b08014
Helpful PU learning wrappers for scikit-learn can be found at: Alexandre Drouin, pu-learning, 2013, https://github.com/aldro61/pu-learning
In addition to our transductive bagging scheme with decision tree base classifiers, we recommend the robust ensemble of support vector machines (RESVM) method introduced by Claesen et al. RESVM is an alternative PU learning method that provides an excellent benchmark. It is implemented here: Marc Claesen, EnsembleSVM, 2014, https://github.com/claesenm/EnsembleSVM and a python wrapper is available here: Marc Claesen, resvm, 2014, https://github.com/claesenm/resvm.
This code is made available under the MIT License.
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size||File type||Python version||Upload date||Hashes|
|Filename, size pumml-0.0.1-py3.6.egg (16.8 kB)||File type Egg||Python version 3.6||Upload date||Hashes View hashes|
|Filename, size pumml-0.0.1-py3-none-any.whl (9.8 kB)||File type Wheel||Python version py3||Upload date||Hashes View hashes|
|Filename, size pumml-0.0.1.tar.gz (8.7 kB)||File type Source||Python version None||Upload date||Hashes View hashes|