Market Sales Prediction by Analyzing Customer Buying Patterns Using Machine Learning

Abstract

To improve a business, a company has to analyze the type of purchases they must keep track of the merchandise that are marketing the foremost in order that they will keep stock of these class merchandise and take away those forms of class that are marketing less. ‘Sales’ is the crucial success issue of a business. Increasing sales may be an excellent impact factor for a developing business. During this trendy time, it may be done by victimizing trendy technology like AI, machine learning, and deep learning. So, we are needing to do that job victimization machine learning by utilizing algorithms. In our research we have a tendency to act on however a mercantile establishment will get a lot of sales from its product victimization of its customer’s previous product shopping for information. We've to preprocess victimization using totally different pre-processing techniques. Information exploration, data transformation and engineering play an important role in predicting correct results. This paper discusses a way to predict sales maximization by information analysis and the way to evaluate the effectiveness of machine learning techniques. Sales analysis of products is one of the major issues of identification buying frequency pattern. We proposed a model to predict seasonal products which is the “ARIMA” model. This model works to do time series analysis. Time series analysis comprises methods for analyzing time series data in order to extract meaningful statistics and other characteristics of the data. Our recommended models can be used to get an idea of which products need to be kept on a shop’s shelves and which products are not for the advantage of the customer. Based on the customer's purchases for a few years this model will be able to recommend which products are more popular in which season.