Differentiable Model Selection for Ensemble Learning

Authors: James Kotary, Vincenzo Di Vito, Ferdinando Fioretto

IJCAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The e2e-CEL training is evaluated on several vision classification tasks: digit classification on MNIST dataset [Deng, 2012], age-range estimation on UTKFace dataset [Zhifei Zhang, 2017], image classification on CIFAR10 dataset [Krizhevsky et al., 2009], and emotion detection on FER2013 dataset [Liu et al., 2016]. Table 2 reports the best accuracy over all the ensemble sizes k of ensembles trained by e2e-CEL along with that of each baseline ensemble model, where each are formed using the same pre-trained base learners.
Researcher Affiliation Academia James Kotary1 , Vincenzo Di Vito1 and Ferdinando Fioretto1 1 University of Virginia {jkotary, vdivitof}@syr.edu, fioretto@virginia.edu
Pseudocode Yes Algorithm 1 summarizes the e2e-CEL procedure for training a selection net. Algorithm 1: Training the Selection Net
Open Source Code No The paper does not contain an explicit statement about releasing the source code for the described methodology, nor does it provide a direct link to a code repository.
Open Datasets Yes digit classification on MNIST dataset [Deng, 2012], age-range estimation on UTKFace dataset [Zhifei Zhang, 2017], image classification on CIFAR10 dataset [Krizhevsky et al., 2009], and emotion detection on FER2013 dataset [Liu et al., 2016].
Dataset Splits Yes In each dataset there is an implied train/test/validation split, so that evaluation of a trained model is always performed on its test portion. Where this distinction is needed, the symbols Xtrain, Xvalid, Xtest are used.
Hardware Specification No The paper does not provide specific details about the hardware (e.g., GPU/CPU models, memory) used to run the experiments, only stating that the selection net uses the 'same CNN architecture as that of the base learner models'.
Software Dependencies No The paper refers to 'standard automatic differentiation employed in machine learning libraries [Paszke et al., 2019]' (which cites PyTorch), but no specific version numbers for any software dependencies or libraries are provided.
Experiment Setup No The paper describes the general approach to training base learners (e.g., specializing on classes) and the selection net's architecture, and mentions `alpha` in Algorithm 1, but it does not provide specific numerical hyperparameter values such as learning rate, batch size, or number of epochs for the main experimental setup.