Global versus cluster-wise regression analyses for prediction of bead geometry in MIG welding process

In this paper, an attempt is made to establish input–output relationships in MIG welding process through regression analyses carried out both globally (i.e., one set of response equations for the entire range of the variables) as well as cluster-wise. It is important to mention that the second approach makes use of the entropy-based fuzzy clusters. The investigation is based on the data collected through full-factorial design of experiments. Results of the above two approaches are compared and some concluding remarks are made. The cluster-wise regression analysis is found to perform a slightly better than the global approach in predicting weld bead-geometric parameters.