Characterization of the Convex Łukasiewicz Fragment for Learning From Constraints

Authors: Francesco Giannini, Michelangelo Diligenti, Marco Gori, Marco Maggini

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The proposed approach is evaluated on a classification task to show how the use of the logical rules can be effective to improve the accuracy of a trained classifier. Section 4 provides an applicative example showing the effect of rules expressed by the proposed fragment in a transductive classification task.
Researcher Affiliation Academia Francesco Giannini, Michelangelo Diligenti, Marco Gori, Marco Maggini Department of Information Engineering and Mathematical Sciences University of Siena Siena, via Roma 56, Italy {fgiannini,diligmic,marco,maggini}@diism.unisi.it
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper does not include an unambiguous statement about releasing code or providing a link to source code for the methodology described.
Open Datasets No The paper mentions using images from the ImageNet database and a benchmark from Winston (Winston and Horn 1986), but it does not provide a direct link, DOI, specific repository name, or proper bibliographic citation to the exact processed dataset used for their experiments.
Dataset Splits No The paper mentions varying the percentage of training supervisions (between 10% and 90%) and evaluating on 'test labels', but it does not explicitly provide specific train/validation/test dataset splits, percentages, or sample counts needed to reproduce the data partitioning.
Hardware Specification No The paper does not provide specific hardware details (e.g., exact GPU/CPU models, processor types with speeds, or memory amounts) used for running its experiments.
Software Dependencies No The paper mentions using a feedforward neural network and Resilient backpropagation, but it does not provide specific software details like library names with version numbers (e.g., Python 3.8, PyTorch 1.9) needed to replicate the experiment.
Experiment Setup Yes A feedforward neural network having one single output neuron and a single hidden layer containing 30 neurons was trained... The single output neuron used a sigmoidal activation function, while the hidden neurons used a rectified linear activation function... trained against the training set labels using a quadratic cost function on the output. Resilient backpropagation... was executed for 500 full-batch iterations and using 0.0001 as initial learning rate for all the weights.