Least Squares Twin Support Vector Machines to Classify End-Point Phosphorus Content in BOF Steelmaking

End-point phosphorus content in steel a basic oxygen furnace (BOF) acts as an indicator of the quality manufactured steel. An undesirable amount is removed from by process dephosphorization. The degree removal captured numerically ‘partition ratio’, given ratio %wt slag and Due to presence multitudes variables, often, it challenging predict partition based on operating conditions. Herein, robust data-driven classification technique least squares twin support vector machines (LSTSVM) applied classify ratio’ two categories (‘High’ ‘Low’) steels indicating greater or lesser removal, respectively. LSTSVM simpler, more robust, faster alternative (TWSVM) with respect non-parallel hyperplanes-based binary classifications. relationship between chemical composition tapping temperatures studied approximately 16,000 heats BOF plants. In our case, relatively higher model accuracy achieved, performed 1.5–167 times than other algorithms.