CUSTOMER RECOMMENDATION USING MACHINE LEARNIGN TECHNIQUES

Abstract

In the customer relationship management business, currently there is a desire for programmed techniques for anticipating future users utilising recommendation engines. There are currently functions for locating "twins," or potential consumers who are similar to current customers, as well as searching the record of clients divided into class. Machine-learning techniques are commonly used in today's recommendation engines. As a result, it's important to figure out which ML algorithms are optimal for building a recommendation engine that can forecast client behaviour. The conditions for determining appropriateness are investigated in this thesis, as well as an evaluation of various off-the-shelf ML techniques. As a result of this, technique of discovering new potential clients, supervised learner models have showed promise. This study is oriented towards learning the behaviour of our customers, so that we can predict out future customer and also can show them their sets of desired products.