Högskolan i Skövde

his.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • apa-cv
  • 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
Root Cause Analysis of Casting Defects using Decision Trees and Oversampling Techniques
University of Skövde, School of Engineering Science.
2021 (English)Independent thesis Advanced level (degree of Master (Two Years)), 12 credits / 18 HE creditsStudent thesis
Abstract [en]

The main goal of this thesis work is to identify the root cause of different types of casting defects in the production of cylinder heads for a Swedish manufacturing company. This research helps the company choose suitable methods to produce with higher performance, remove casting defects, and improve their efficiency. This work will involve developing and implementing interpretable Machine Learning (ML) and Rule-Based approaches to Root Cause Analysis (RCA). Statistical analysis of the casting data is required to understand and identify the distribution of casting defects and process parameters. This analysis will enable the detection of outliers and errors. Standard data imputation methods will fill in the missing parameter values. Several oversampling methods, such as SMOTE and ADASYN, are implemented in this research. Comparing the effectiveness of these methods on casting data is essential to fix an imbalanced dataset. Machine learning models such as Random Forest (RF) will be trained to classify the processed data. Finally, interpretable rules that explain the root causes for various defect types must be extracted from the rule extraction techniques.

Place, publisher, year, edition, pages
2021. , p. 83
Keywords [en]
Root Cause Analysis, Casting Defects, Over-sampling, SMOTE, Imbalanced Data, Machine Learning, Classification, Random Forest
National Category
Production Engineering, Human Work Science and Ergonomics
Identifiers
URN: urn:nbn:se:his:diva-24624OAI: oai:DiVA.org:his-24624DiVA, id: diva2:1906562
Subject / course
Virtual Product Realization
Supervisors
Examiners
Available from: 2024-10-17 Created: 2024-10-17 Last updated: 2025-09-29Bibliographically approved

Open Access in DiVA

fulltext(5294 kB)842 downloads
File information
File name FULLTEXT01.pdfFile size 5294 kBChecksum SHA-512
da374c96bb179824fec5cc20e92dc0feb29475f10c062b04a96881f02a3a280d4199ef6ff35e2fcceb99a75d783c1aa5d016a37a7018a41dbf0dd67a5989519f
Type fulltextMimetype application/pdf

By organisation
School of Engineering Science
Production Engineering, Human Work Science and Ergonomics

Search outside of DiVA

GoogleGoogle Scholar
Total: 845 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: 345 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • apa-cv
  • 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