sklvq: Scikit Learning Vector Quantization

Authors: Rick van Veen, Michael Biehl, Gert-Jan de Vries

JMLR 2021 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical The theory behind this design is described in this paper. To facilitate adoptions and inspire future contributions, sklvq is publicly available on Github (under the BSD license) and can be installed through the Python package index (Py PI). Next to being well-covered by automated testing to ensure code quality, it is accompanied by detailed online documentation. The documentation covers usage examples and provides an in-depth API including theory and scientific references. Keywords: Python, scikit-learn, learning vector quantization, matrix relevance learning... Here we discuss the theoretical concepts and link them with the reasoning behind the implementation in the code repository... In the previous section, we have shown the theory behind the design and implementation of sklvq and the resulting advantages. This section, particularly Table 1, provides a comparison of the resulting functionality with that of other LVQ toolboxes.
Researcher Affiliation Collaboration Rick van Veena EMAIL Michael Biehla,b EMAIL a Bernoulli Institute for Mathematics, Computer Science and Artificial Intelligence, University of Groningen, The Netherlands, b SMQB, Institute of Metabolism and Systems Research, College of Medical and Dental Sciences, Birmingham, UK Gert-Jan de Vries EMAIL Philips Research, Healthcare, Eindhoven, The Netherlands
Pseudocode No The paper provides mathematical equations describing the objective function and gradient updates (e.g., Equation 1, 2, 3), and discusses algorithm components. However, it does not include any clearly labeled pseudocode or algorithm blocks with structured steps.
Open Source Code Yes To facilitate adoptions and inspire future contributions, sklvq is publicly available on Github (under the BSD license) and can be installed through the Python package index (Py PI). Next to being well-covered by automated testing to ensure code quality, it is accompanied by detailed online documentation. The documentation covers usage examples and provides an in-depth API including theory and scientific references. 1. https://github.com/rickvanveen/sklvq/releases/0.1.2
Open Datasets No The paper discusses the sklvq software package for Learning Vector Quantization algorithms but does not conduct experiments on a specific, publicly available dataset. It defines a generic dataset 'D = ( xi, yi) | xi RN, yi {1, . . . , C} P i=1' for theoretical explanation, but no concrete access information to any dataset is provided.
Dataset Splits No The paper describes a software package and its design, rather than conducting experiments on a specific dataset. Therefore, there is no information about training/test/validation dataset splits.
Hardware Specification No The paper focuses on the design and implementation of a software package. It does not describe any specific hardware (like GPU or CPU models) used for running experiments, as the paper itself does not report experimental results.
Software Dependencies No The paper mentions 'Python' as the implementation language and 'scikit-learn (Pedregosa et al., 2011) compatible' for the sklvq package. However, it does not provide specific version numbers for Python, scikit-learn, or any other software dependencies, which is required for reproducibility.
Experiment Setup No The paper introduces the sklvq software package and describes its modular design and theoretical underpinnings. It does not present specific experimental results with concrete hyperparameters, training configurations, or system-level settings. The mention of 'step size η(t)' is part of the theoretical algorithm description, not an experimental setup detail.