Abstract:
Customer plays a crucial role for the success of any businesses. It is difficult to survive in
the competitive business environment without attracting the customers. Since, most successful
businesses continuously do the research and development part for their customers in the fields
of customer identification, customer attraction, customer retention, and customer development
to achieve a high level of customer relationship. All the competitors are trying to make more
customer profit to survive in their business. To achieve this, businesses required to increase
their capabilities on understanding customer behavioural patterns and preferences. The major
aim of this study is to categorize the customers based on behaviours and develop a prediction model to predict future customer categories. The current study is carried under the two
phases. Initial phase is focusing on clustering the customers based on their behaviours by using K-means++ algorithm. Recency, Frequency and Monetary (RFM) attributes are used to
cluster the customers and the second phase is to develop the Artificial Neural Network (ANN)
model to predict future customer’s category based on their usage behaviours. Dataset consists
of 5,000 customer details in a particular business with RFM attributes to cluster, train and test
the model. K-means++ algorithm used to cluster and the final weighted cluster centroids are
calculated based on the weights of RFM attributes. Target attributes are generated by analysing
the final weighted cluster centroids for the ANN model. Confusion matrix used to evaluate the
performance of ANN model. Existing researches applied unsupervised or supervised learning
algorithms separately. But this research study integrates both. Therefore, this model well fit for
any business industries without any errors.
