Axiomatic Characterization of Data-Driven Influence Measures for Classification

Authors: Jakub Sliwinski, Martin Strobel, Yair Zick718-725

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

Reproducibility Variable Result LLM Response
Research Type Experimental 7 Experimental Results In what follows, we apply MIM, Parzen and a version of LIME on a facial expression dataset. We ran our experiments using a workstation with a quad core Intel i7 CPU, and 16GB of RAM.
Researcher Affiliation Academia Jakub Sliwinski ETH Zurich Zurich, Switzerland jsliwinski@ethz.ch Martin Strobel National University of Singapore Singapore mstrobel@comp.nus.edu.sg Yair Zick National University of Singapore Singapore zick@comp.nus.edu.sg
Pseudocode No No pseudocode or algorithm blocks were found in the paper. The paper focuses on theoretical derivations and proofs.
Open Source Code No The paper refers to a 'full version of this paper (Sliwinski, Strobel, and Zick 2018)' but does not explicitly state that the source code for the methodology presented in this paper is open-source or provide a link.
Open Datasets Yes The dataset used for this experiment is a part of the Facial Expression Recognition 2013 dataset (Goodfellow et al. 2013).
Dataset Splits No The paper mentions the dataset size and composition ('12 156 48 48 pixel grayscale images of faces, evenly divided between happy and sad facial expressions') but does not provide specific training, validation, or test dataset splits.
Hardware Specification Yes We ran our experiments using a workstation with a quad core Intel i7 CPU, and 16GB of RAM.
Software Dependencies No The paper mentions using MIM, Parzen, and LIME but does not provide specific version numbers for any software dependencies or libraries.
Experiment Setup Yes A parametric Parzen influence measure with σ = 4.7 and a monotone influence measure with α(d) = 1 d2 were run on some of the images. Further, we used a black-box data version of LIME as described in detail in the full version of this work (Sliwinski, Strobel, and Zick 2018). For the α parameter in Equation 7, we choose αρ(d) = exp( d2/ρ2) with ρ = 3 as a Kernel function.