Theano implementation of Logistic Disjunctive Normal Networks.
Choose between two datasets of noisy_moons
or noisy_circles
by choosing data, labels = create_data(dataset='moons')
or create_data(dataset='circles')
.
Choose between product approximation or min_max approximation of logical gates by choosing ldnn.classify(X, approx='product')
or approx='min_max'
Sajjadi, Mehdi, Mojtaba Seyedhosseini, and Tolga Tasdizen. "Disjunctive Normal Networks." arXiv preprint arXiv:1412.8534 (2014).
http://arxiv.org/abs/1412.8534
This code is distributed without any warranty, express or implied.