Multi-criteria decision-making for purchasing cell phones using machine learning approach

Abstract

The process of selecting and purchasing cell phones is a multi-criteria decision-making (MCDM) problem with conflicting and diverse objectives. This study discusses various techniques involved in selecting and purchasing a cell phone by using machine learning approach. The responses of the participants were sought through a questionnaire which gave them different options with regard to the latest features available in a cell phone. Seven independent input variables – cost, battery backup, rear camera, weight, size, memory and operating system, were provided to the participants to elicit their responses. Each of the input variables was measured on a scale expressed in linguistic terms as low, medium and high. Mamdani approach, traditional fuzzy reasoning tool (FLC) and neuro-fuzzy system (ANFIS) were used to design three input and one output processes. The back-propagation algorithm formed the basis for application of the neuro-fuzzy system. Two traditional fuzzy reasoning tools – the artificial neural network (ANN) approach and the neuro-fuzzy system, were used to arrive at more accurate understanding of the process of selecting a cell phone for personal use.