https://www.mdu.se/

mdu.sePublications
System disruptions
We are currently experiencing disruptions on the search portals due to high traffic. We are working to resolve the issue, you may temporarily encounter an error message.
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
SUPERVISED LEARNING TECHNIQUES FORANOMALY DETECTION ANDCLASSIFICATION IN A MODULARMANUFACTURING ENVIRONMENT
Mälardalen University, School of Innovation, Design and Engineering.
Mälardalen University, School of Innovation, Design and Engineering.
2022 (English)Independent thesis Advanced level (degree of Master (Two Years)), 80 credits / 120 HE creditsStudent thesis
Abstract [en]

An anomaly is an event or data pattern that differs from the expected behavior. Anomaly de-tection is an approach that aims to detect anomalous data points, and machine learning techniquesare one of the most widely used approaches for anomaly detection. This thesis focuses on anomalydetection and classification in a modular manufacturing environment, using existing software thatsimulates an ice cream factory. Existing software contains different modules and those modules arecontrolled using controller and orchestrator. The existing software is upgraded with a new modulethat enables the injection of anomalies into the simulation process. During the development wefollowed software engineering practices for software design and development. A new dataset withnormal and anomalous behavior of the manufacturing process is generated and different supervisedmachine learning algorithms are used to detect and classify anomalies. Our objective is to generatea dataset that is completely balanced. (Additionally, we created our dataset which is composed ofthe different files). Those files contain normal behavior and abnormal behavior. Algorithms thatwere evaluated include Decision Tree, Random Forest, K-Nearest Neighbour, and Artificial NeuralNetwork. The best accuracy was achieved with the Decision Tree (without normalization) of 96%for multi-class classification problems. K-Nearest Neighbour shows the best performance for multi-class classification, resulting in 89%, and 90% for binary classification. Artificial Neural Networkgave us 91% for anomaly detection and 85% for multi-class classification. Also, Random Forestobtained 61% for binary classification and 88% for multi-class problems. Additionally, we test theperformance mode of DT with normalized data and we obtained 97% for binary classification and73% for multi-class classification.i   

Place, publisher, year, edition, pages
2022. , p. 42
National Category
Engineering and Technology
Identifiers
URN: urn:nbn:se:mdh:diva-61532OAI: oai:DiVA.org:mdh-61532DiVA, id: diva2:1728306
Subject / course
Computer Science
Supervisors
Examiners
Available from: 2023-01-19 Created: 2023-01-18 Last updated: 2023-01-19Bibliographically approved

Open Access in DiVA

fulltext(3283 kB)187 downloads
File information
File name FULLTEXT01.pdfFile size 3283 kBChecksum SHA-512
e6fd3ce4483e661845febc1066e928f4f0de840376b90226ac8d78647bf3536bd0dac57b7a374b5a7e484e0033e7e2c66f56cf7191c24d496108c8a031fb0563
Type fulltextMimetype application/pdf

By organisation
School of Innovation, Design and Engineering
Engineering and Technology

Search outside of DiVA

GoogleGoogle Scholar
Total: 187 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: 263 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