liu.seSearch for publications in DiVA
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • 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
Naive Bayes-klassificering av förarbeteende
Linköping University, Department of Computer and Information Science, Software and Systems.
Linköping University, Department of Computer and Information Science, Software and Systems.
2017 (Swedish)Independent thesis Basic level (degree of Bachelor), 10,5 credits / 16 HE creditsStudent thesisAlternative title
Classifying driving behaviour using Naive Bayes (English)
Abstract [sv]

Att kunna klassificera en körstil implicerar klassificering av körbeteende, vilket ligger till grunden för miljö- och säkerhetsklassificering för körningar.

I det här arbetet har vi låtit två förare köra en bil med en förhoppning att kunna klassificera vem det var som körde bilen. Målet var att kunna förutspå föraren med en korrekthet på 80-90% givet endast hastighet samt varvtal som samlas genom ODB:II-porten via CAN-bussen i fordonet. Angreppsättet på detta arbete liknar det för textklassificering, nämligen att använda två vanliga klassificeringsmetoder från just textklassificering — Multinominal och Gaussisk Naive Bayes tillsammans med N-gram samt diskretisering.

Vi fann genom att använda Multinominal Naive Bayes med 4-gram samt icke-diskretiserade respektive diskretiserade hastighet- och varvtalsvärden kunde klassificera förare med 91.48% korrekthet. 

Abstract [en]

To be able to classify a driving style implies that you classify a driving behaviour, which is the foundation of safety and environmental driving classification.

In this thesis we have let two drivers drive a car in an attempt to classify, with a desired accuracy of around 90%, which one of us drove the car. This was done by exclusively using speed and rpm data values provided from the OBD:II port of the car via the CAN-bus. We approched this problem like you would a text classification one, thus using two common models of Naive Bayes — Multinominal and Gaussian Naive Bayes together with N-gram and discretization.

We found that using Multinominal Naive Bayes consisting of 4-gram resulted in an avarage accuracy of 91.48% in predicting the driver, non-discretized speed and discretized rpm values.

Place, publisher, year, edition, pages
2017. , p. 34
National Category
Computer Sciences
Identifiers
URN: urn:nbn:se:liu:diva-139957ISRN: LIU-IDA/LITH-EX-G--17/070--SEOAI: oai:DiVA.org:liu-139957DiVA, id: diva2:1135249
Subject / course
Computer science
Presentation
2017-06-22, Alan Turing, Linköping, 08:15 (Swedish)
Supervisors
Examiners
Available from: 2017-08-24 Created: 2017-08-22 Last updated: 2018-01-13Bibliographically approved

Open Access in DiVA

fulltext(2186 kB)35 downloads
File information
File name FULLTEXT01.pdfFile size 2186 kBChecksum SHA-512
b6a6ba490beae3006bf46ca2bd299a1590cc16970e48c363a94f012d3cdf3526f9ad8763d7208c19003c9c893e6130e1ff096e0a02f580150e5989ed87c14c3f
Type fulltextMimetype application/pdf

Search in DiVA

By author/editor
Henrik, ThoresonRobin, Wesslund
By organisation
Software and Systems
Computer Sciences

Search outside of DiVA

GoogleGoogle Scholar
Total: 35 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 267 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • 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