Deep Gamblers: Learning to Abstain with Portfolio Theory

Authors: Ziyin Liu, Zhikang Wang, Paul Pu Liang, Russ R. Salakhutdinov, Louis-Philippe Morency, Masahito Ueda

NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments show that our method can identify uncertainty in data points, and achieves strong results on SVHN and CIFAR10 at various coverages of the data.
Researcher Affiliation Academia Institute for Physics of Intelligence & Department of Physics, University of Tokyo Machine Learning Department, Carnegie Mellon University Language Technologies Institute, Carnegie Mellon University {zliu,wang}@cat.phys.s.u-tokyo.ac.jp ueda@phys.s.u-tokyo.ac.jp {pliang,rsalakhu,morency}@cs.cmu.edu
Pseudocode No The paper does not contain any pseudocode or clearly labeled algorithm blocks.
Open Source Code No The paper does not provide any statement about releasing its source code or a link to a code repository.
Open Datasets Yes Experiments show that our method can identify uncertainty in data points, and achieves strong results on SVHN and CIFAR10 at various coverages of the data. ... SVHN [31] (Table 3), CIFAR10 [23] (Table 4) and Cat vs. Dog (Table 5).
Dataset Splits Yes The best models of ours for a given coverage are chosen using a validation set, which is separated from the test set by a fixed random seed, and the best single model is chosen by using the model that achieves overall best validation accuracy.
Hardware Specification No The paper does not provide any specific details about the hardware used to run the experiments, such as GPU models, CPU types, or cloud instance specifications.
Software Dependencies No The paper mentions using a 'version of VGG16' but does not specify any software dependencies with version numbers (e.g., programming languages, libraries, frameworks).
Experiment Setup Yes We use a version of VGG16 that is especially optimized for small datasets [27] with batchnorm and dropout. ... A grid search is done over hyperparameter o with a step size of 0.2. ... we train a network with 2 hidden layers each with 50 neurons and tanh activation. ... The model is a simple 4-layer CNN.