Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

Learning vector quantization for multiclass classification: application to characterization of plastics.

Gavin R Lloyd¹, Richard G Brereton, Rita Faria

¹Centre for Chemometrics, School of Chemistry, University of Bristol Cantocks Close, Bristol BS8 1TS, United Kingdom.

Journal of Chemical Information and Modeling

|July 5, 2007

Summary

This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Regional HNSCC metabolomics reveals widespread changes to one-carbon metabolism and S-adenosylmethionine metabolism across tumour core, tumour edge and adjacent non-tumour tissues.

British journal of cancer·2026

Same author

A burning mystery: one artery to rule them all?

Netherlands heart journal : monthly journal of the Netherlands Society of Cardiology and the Netherlands Heart Foundation·2026

Same author

Modelling the cost-effectiveness of interventions to treat or prevent neuropathic ulcers arising from leprosy: application to L-PRF.

Cost effectiveness and resource allocation : C/E·2026

Same author

Association of high-risk CT coronary artery plaque features with major adverse cardiovascular events: a prespecified secondary analysis of the DISCHARGE trial.

European radiology·2026

Same author

Endocrine and metabolic determinants of cardiometabolic risk in mild autonomous cortisol secretion.

EBioMedicine·2026

Same author

Revised methods guide for economic evaluation studies of health technologies in Portugal.

International journal of technology assessment in health care·2026

Same journal

Advancing Biochemical Molecule Registration, Representation and Search for New Drug Modalities.

Journal of chemical information and modeling·2026

Same journal

A Unified Molecular Graph and Protein Language Model Framework for Predicting Human Drug-Hormone Receptor Interactions with Structure-Aware Validation.

Journal of chemical information and modeling·2026

Same journal

Intricate Role of Cholesterol in Membrane Fusion.

Journal of chemical information and modeling·2026

Same journal

tmGNN-XAI: An Explainable Graph Neural Network Tool for Predicting Electronic Properties of Transition Metal Complexes from SMILES.

Journal of chemical information and modeling·2026

Same journal

QSAR in the Browser: An Interactive Cheminformatics Web Application.

Journal of chemical information and modeling·2026

Same journal

FoldDoF: Utilizing the Primary Degrees of Freedom of Protein Backbone for Geometric Modeling and Generation.

Journal of chemical information and modeling·2026

See all related articles

Learning Vector Quantization (LVQ) effectively classifies polymer thermal properties, outperforming Mahalanobis distance. This neural network approach models multiple classes, offering a robust alternative for complex datasets.

Area of Science:

Materials Science
Computer Science
Machine Learning

Background:

Traditional boundary methods like Support Vector Machines struggle with multi-class problems.
Learning Vector Quantization (LVQ) offers a neural network-based approach for multi-class classification.
LVQ creates codebook vectors for each class to define decision boundaries.

Purpose of the Study:

To evaluate the performance of LVQ algorithms (LVQ1 and LVQ3) for classifying commercial polymers based on thermal properties.
To compare LVQ's effectiveness against the Mahalanobis distance method for multi-class polymer classification.
To analyze classification accuracy and misclassifications using confusion matrices.

Main Methods:

Detailed description of LVQ1 and LVQ3 algorithms.

Related Experiment Videos

Application of LVQ to a dataset of thermal properties for 293 commercial polymers across nine classes.

Comparative analysis with the Mahalanobis distance method.

Validation using iterative data splits into training and testing sets.

Main Results:

LVQ demonstrated superior performance compared to the Mahalanobis distance method for the tested polymer dataset.
LVQ's advantage lies in its ability to model many classes without assumptions on data distribution, unlike Mahalanobis distance.
Confusion matrices revealed misclassification patterns interpretable by the chemical similarity of polymer grades.

Conclusions:

LVQ is a highly effective method for classifying materials based on complex property datasets.
LVQ provides a more flexible and accurate approach than Mahalanobis distance for non-ellipsoidal data distributions.
The study highlights LVQ's potential in materials science for nuanced classification tasks.