keras-quadopt: Solving Constrained Quadratic Optimization Problems with Reverse Automatic Differentation (Keras/TF2)

Solving quadratic optimization problems with resource constraints and upper boundaries using TF2/Keras.

Usage

import tensorflow as tf
import keras_quadopt as kqp
import time

# goodness scores
good = tf.constant([.51, .53, .55, .57])

# similarity matrices
simi_1 = tf.constant([
    [1, .9, .8, .7],
    [.9, 1, .6, .5],
    [.8, .6, 1, .4],
    [.7, .5, .4, 1],
])

simi_2 = tf.constant([
    [1, .7, .8, .3],
    [.7, 1, .4, .2],
    [.8, .4, 1, .6],
    [.3, .2, .6, 1],
])

# preference parameters
lam = 0.4
beta_1 = 0.25
beta_2 = 0.75

# compute weights
simi = kqp.aggregate_matrices(simi_1, beta_1, simi_2, beta_2)

start = time.time()
wbest, fbest = kqp.get_weights(good, simi, lam)
print(f"elapsed: {time.time() - start}")

Appendix

Installation

The keras-quadopt git repo is available as PyPi package

pip install keras-quadopt
pip install git+ssh://[email protected]/ulf1/keras-quadopt.git

Install a virtual environment

python3 -m venv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install -r requirements.txt --no-cache-dir
pip install -r requirements-dev.txt --no-cache-dir

(If your git repo is stored in a folder with whitespaces, then don't use the subfolder .venv. Use an absolute path without whitespaces.)

Python commands

• Check syntax: flake8 --ignore=F401 --exclude=$(grep -v '^#' .gitignore | xargs | sed -e 's/ /,/g')
• Run Unit Tests: PYTHONPATH=. pytest

Publish

python setup.py sdist 
twine upload -r pypi dist/*

Clean up

find . -type f -name "*.pyc" | xargs rm
find . -type d -name "__pycache__" | xargs rm -r
rm -r .pytest_cache
rm -r .venv

Support

Please open an issue for support.

Contributing

Please contribute using Github Flow. Create a branch, add commits, and open a pull request.

Acknowledgements

The "Evidence" project was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) - 433249742 (GU 798/27-1; GE 1119/11-1).

Maintenance

• till 31.Aug.2023 (v0.1.0) the code repository was maintained within the DFG project 433249742
• since 01.Sep.2023 (v0.2.0) the code repository is maintained by Ulf Hamster.