A Survey on Customer Segmentation using Machine Learning Algorithms to Find Prospective Clients

Numerical classification has occupied an important position in economic development due to the rapid growth of the world's capital and society. The focus is to enhance strategic governance by improving some common virtues regarding service quality systems. Market segmentation not only grasps the customer's geographic, demographic, psychographic and behavioral traits but also governs the market according to the consumer's needs, by undertaking the correlating procedure to establish a better retailer-consumer relationship. This can be achieved by a concept that comes under unsupervised learning, known as cluster analysis. In this method, customers are divided into ‘n’ number of groups based upon the above-mentioned traits, clustering means grouping the information based on similarities in the dataset. Customers belonging to a particular group have some common traits. Customers are grouped in a way that a customer belonging to a particular group shares a common interest with other customers of the same group. In this paper, we have studied and explained various algorithms related to clustering which help segment customers according to their needs. With respect to their datasets, these algorithms are analyzed and compared using metrics like Silhouette and Davies Bouldin.