ABC classification using MCDA, DDF and TOPSIS approach

This paper outlines the way of classifying stocking items using multi-criteria decision analysis into A, B and C classes. A combination of directional distance function and technique for order of preference by similarity to ideal solution are suggested in this regard to select the pair of ideal points. The prescribed model argues in favour of designing two frontiers to aid in the categorisation of items with an underlying assumption of convexity. Moreover, the proposed approach has a guaranteed solution in the presence of negative data where the conventional data envelopment analysis models find difficulties in measuring the convex efficiency scores for the given set of alternatives. The two frontier approach is meant for enclosing each alternative inside the worst and best frontiers. Unlike the traditional data envelopment analysis models two convex efficiency scores are computed. Inefficiency scores deduced from these two frontiers are then used for measuring a proximity score which ultimately discriminates the items into three categories.