Asia Pacific University Library catalogue

A thesis submitted in fulfillment of the requirements for the award of the degree of MSc. in Data Science & Business Analytics (UCMF1808DSBA)

The sales prediction performs an important role in every market process and the decision making for sales of products are important to follow the demand of consumers. This research focuses on the study of sales prediction and business decision making for e-commerce retail business with data science research approach. The main goal of research is to perform data science analysis on collected sales data with minimum predictor variables to create advance sales prediction model to forecast future sales performance and identify new business decisions depending on the predicted results and data analysis report. Applying business intelligence, statistical modelling and machine learning are proposed to create intelligent prediction and analytical solution of retail business and optimization methods are also applied to create business decisions. Business decisions are made depending on analysis result using mathematical programming methods to set profitable decisions which can minimize or maximize capital and control the stock with decision results. The research mainly depends on the quantitative study in which statistical modelling techniques are applied using different predictive models and results gained are also in quantitative form. The result of this research is solving additional sales problems in online retail environment using predictable machine learning, statistical data analysis, business intelligence and applied operational research. The case study, methodology and decision are made from point of data science and business analyst view.