University of Twente Student Theses
Using Machine Learning to Improve the Cost Estimation Process at a Versatile Manufacturing Company
Vlastuin, A. van (2023) Using Machine Learning to Improve the Cost Estimation Process at a Versatile Manufacturing Company.
|
PDF
29MB |
| Abstract: | This thesis examines the feasibility of machine learning models to estimate the (engineering) costs for a versatile manufacturing company (VMC). To deal with the speed, accuracy and consistency requirements of VMCs’ quotation processes, the company identified a need for a time-efficient, objective and interpretable cost estimation method. Drawing on interviews and existing literature on cost estimation and design & engineering costs, this study identifies several cost drivers. We focus on regression analysis to construct a model that is understandable and implementable by the company, and optimise four variants. The analysis shows that the Lasso Regression model performs best; slightly better than the company’s manual method based on the mean squared error, but with a 57% higher percent error. The model uses only one feature, indicating that many of the engineered features offered little predictive value. Therefore, the thesis concludes that our model is not accurate enough to offer a reliable improvement over the old method. Several recommendations are offered for improving the data quality and reliability, for the company and for other VMCs like it. |
| Item Type: | Essay (Bachelor) |
| Faculty: | BMS: Behavioural, Management and Social Sciences |
| Subject: | 54 computer science |
| Programme: | Industrial Engineering and Management BSc (56994) |
| Link to this item: | https://purl.utwente.nl/essays/95531 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page