A benchmark comparison of Gaussian process regression, support vector machines, and ANFIS for man-hour prediction in power transformers manufacturing

Işık K., Alptekin S. E.

26th International Conference on Knowledge-Based and Intelligent Information & Engineering Systems, Verona, İtalya, 7 - 09 Eylül 2022, cilt.207, ss.2567-2577

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası: 207
  • Doi Numarası: 10.1016/j.procs.2022.09.315
  • Basıldığı Şehir: Verona
  • Basıldığı Ülke: İtalya
  • Sayfa Sayıları: ss.2567-2577
  • Galatasaray Üniversitesi Adresli: Evet

Özet

Production times affect the product valuations because they are directly relevant to a product's cost. The man-hour unit is widely used for measuring the production times, especially in labor-intensive manufacturing environments, and it is considered when making product valuations before the tendering phase of power transformer projects. To our knowledge, no study has been made to improve the prediction performance of man-hour estimations in this field. In this study, SVM, GPR, and ANFIS models were simultaneously applied to predict man-hours in Power Transformer manufacturing.