On the veracity of local, model-agnostic explanations in audio classification: targeted investigations with adversarial examples

V Praher, K Prinz, A Flexer, G Widmer - arXiv preprint arXiv:2107.09045, 2021 - arxiv.org
Local explanation methods such as LIME have become popular in MIR as tools for
generating post-hoc, model-agnostic explanations of a model's classification decisions. The …

Tracing back music emotion predictions to sound sources and intuitive perceptual qualities

S Chowdhury, V Praher, G Widmer - arXiv preprint arXiv:2106.07787, 2021 - arxiv.org
Music emotion recognition is an important task in MIR (Music Information Retrieval)
research. Owing to factors like the subjective nature of the task and the variation of …

Reliable local explanations for machine listening

S Mishra, E Benetos, BLT Sturm… - 2020 International Joint …, 2020 - ieeexplore.ieee.org
One way to analyse the behaviour of machine learning models is through local explanations
that highlight input features that maximally influence model predictions. Sensitivity analysis …

Towards Musically Meaningful Explanations Using Source Separation

V Haunschmid, E Manilow, G Widmer - arXiv preprint arXiv:2009.02051, 2020 - arxiv.org
Deep neural networks (DNNs) are successfully applied in a wide variety of music
information retrieval (MIR) tasks. Such models are usually considered" black boxes" …

A Critical Look at the Music Classification Experiment Pipeline: Using Interventions to Detect and Account for Confounding Effects

F Rodríguez-Algarra - 2020 - qmro.qmul.ac.uk
This dissertation focuses on the problemof confounding in the design and analysis of music
classification experiments. Classification experiments dominate evaluation of music content …

Behavior Action Mining

P Su, D Zeng, H Zhao - IEEE Access, 2019 - ieeexplore.ieee.org
The actionable behavioral rules suggest specific actions that may influence certain behavior
in the stakeholders' best interest. In mining such rules, it was assumed previously that all …