Development and validation of a hybridised algorithm involving AHP and machine learning for automobile vehicle selection

The problem of selecting an automobile has always been one of the most complex decisions to make, given a person's social and economic life. It is often resolved either through a qualitative judgement of vehicles or through multiple criteria decision-making (MCDM) methods in an algorithmic way. However, the modern machine learning (ML) procedures have surfaced themselves as efficient techniques in the field of recommendation engines (REs) to predict the items that may be useful to the customers according to their preferences. In this paper, an attempt has been made to study the automobile vehicle selection (AVS) problem in an innovative manner by hybridising the analytic hierarchical process (AHP) with the collaborative filtering (CF) technique to construct a selector to recommend the customers precisely one pair of cars that would suit best to their preference. The proposed algorithm provides an efficient way to map the satisfaction level of the customers by eliminating the vagueness and complexity in the selection process. We have validated the algorithm using real-life datasets collected by administering an exploratory survey across geographies, including India.