Finding ideal geographical location for businesses using machine learning technique

Abstract

The world has become a place where the economy is at the epicenter of it all. World economic growth has paved the way for people to enrich their lives with all sorts of blessings. A major chunk of this shift in the world's treasury all comes from the tireless endeavors of a uent and ourishing businesses. The more a business thrives, the more economic sustainability it brings upon the society. And one of the key factors of building up a thriving business is what motivated us to forgo on our research. Location analytics in recent times plays an important role in making a sustainable and pro table business. Very often trade and commerce rely on uninformed struggles to analyze the perfect location for their establishment. Hence, we used unsupervised learning to evaluate a dataset and create a decision making model to accurately investigate whether a location will be be tting for a particular business model based on customer behaviour and interests. As such we built a dataset concentrating on questionnaire responses taken via online survey and tested our model on the gathered data. We exercised the analyzed dataset in di erent clustering algorithms such as kmeans, minibatch kmeans and hierarchical clustering. Finally, using a decision tree model, we were able to extract an explanatory rule in terms of quantitative values, which were further discerned in making an elaborate assumption. Hence, based on consumer habits and interests, we concluded that we could analyze which location or marketplace would be better suited for which accessories a seller is selling, and hence suggest the perfect location for his business to thrive upon.