To Örebro University

oru.seÖrebro University Publications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Gas Discrimination with One-class Support Vector Machines
Örebro University, School of Science and Technology.
2021 (English)Independent thesis Basic level (professional degree), 10 credits / 15 HE creditsStudent thesis
Abstract [en]

This is a bachelor thesis about investigating the viability of using a one-class support vector machine (OCSVM) to detect a target substance in a gas regardless of inference, using partially selective metal-oxide (MOX) gas sensors.Many applications of an electronic nose (e-nose) require the sensor to be used in complex scenarios, where environmental conditions are uncontrolled and unknown non-target substances might be present. The latter hinders the use of multiclass classifiers but leaves one-class classifiers as a possible option. If using one class classification to interpret response patterns from an array of metal-oxide gas sensors for gas discrimination tasks prove viable, this can potentially broaden the applicability of low-cost MOX sensors-based e-noses in real-world scenarios.Throughout this investigation, several OCSVM-models have been trained and tested using data collected from practical sampling of several gas classes comprised of different mixes of the same two analytes.Four different training sets will be utilized for model training. One of these sets will contain data generated exclusively from samples taken on pure target analyte with no inference present, while the others are comprised of data collected from a mixture of target analyte and inference analyte.The OCSVM models trained on these different training sets are tuned in terms of functional parameters, and then the tuned models are used to perform gas classification on test sets that include pure analytes or gas mixtures. The model implemented relies on the scikit-learn library, version 0.23.2.Based on the gas classification results, OCSVM models investigated in this work are validated for the use of one-class classification for gas discrimination tasks.

Place, publisher, year, edition, pages
2021. , p. 83
National Category
Computer Sciences
Identifiers
URN: urn:nbn:se:oru:diva-95313OAI: oai:DiVA.org:oru-95313DiVA, id: diva2:1608348
Subject / course
Computer Engineering
Supervisors
Examiners
Available from: 2021-11-03 Created: 2021-11-03 Last updated: 2021-11-03Bibliographically approved

Open Access in DiVA

No full text in DiVA

By organisation
School of Science and Technology
Computer Sciences

Search outside of DiVA

GoogleGoogle Scholar

urn-nbn

Altmetric score

urn-nbn
Total: 96 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf