In 2003.

abstract bibtex

abstract bibtex

Statistical models that predict the tensile strength of low-alloyed steel plates using the element concentrations and some variables of the rolling process were developed. The purpose of the work was to develop a new predicting model for Rautaruukki’s steel plate mill. The model will be used mainly in the product design of steel plates. The standard deviation of the error term of the best regression model was 10 MPa, which can be considered very good. The performance of the regression model was compared to a neural network model, but significantly better predictions were not achieved with neural networks than with regression models. The quantity of data used was very large, and special attention was therefore paid to avoid overfitting.

@inProceedings{ title = {Modelling the Strength of Steel Plates Using Regression Analysis and Neural Networks}, type = {inProceedings}, year = {2003}, id = {6ecfc171-dba7-3b9f-8ec6-cdcdfca7e0c2}, created = {2019-11-19T13:01:03.739Z}, file_attached = {false}, profile_id = {bddcf02d-403b-3b06-9def-6d15cc293e20}, group_id = {17585b85-df99-3a34-98c2-c73e593397d7}, last_modified = {2019-11-19T13:46:10.222Z}, read = {false}, starred = {false}, authored = {false}, confirmed = {true}, hidden = {false}, citation_key = {isg:458}, source_type = {inproceedings}, notes = {Proceedings of International Conference on Computational Intelligence for Modelling, Control and Automation (CIMCA'2003)}, private_publication = {false}, abstract = {Statistical models that predict the tensile strength of low-alloyed steel plates using the element concentrations and some variables of the rolling process were developed. The purpose of the work was to develop a new predicting model for Rautaruukki’s steel plate mill. The model will be used mainly in the product design of steel plates. The standard deviation of the error term of the best regression model was 10 MPa, which can be considered very good. The performance of the regression model was compared to a neural network model, but significantly better predictions were not achieved with neural networks than with regression models. The quantity of data used was very large, and special attention was therefore paid to avoid overfitting.}, bibtype = {inProceedings}, author = {Juutilainen I Röning J, Myllykoski L} }

Downloads: 0