ABC classification using extended R-model, SVM and Lorenz curve

ABC analysis has wide applications in the domain of management. Myriads of innovative models have been put forward in this context. Unfortunately, for a dataset, these methods could only reveal diverse solutions. Also, hardly any investigation stressed the extent of inequality obtained from any classification. This paper extends the linear model proposed by Ramanathan (Comput Oper Res 33(3):695-700, 2006) to delineate a non-linear model embedded with a Support Vector Machine to derive the weights needed for each class from the training data. A rank-based approach is preferred to classify the testing data. Lastly, to compare the extent of inequality evoked from the output of the model, an Ideal Lorenz curve is developed as a benchmark. Chiefly, this framework controls the cumulative percentage scores against the cumulative percentage population size of the elements within each group to get a better inequality.