Mining sales data to identify customer profiles, and predict sales

Lovas, Lukas (2024) Mining sales data to identify customer profiles, and predict sales. Bachelor's Thesis, Computing Science.

Preview	Text BT.pdf Download (2MB) | Preview
	Text toestemming.pdf Restricted to Registered users only Download (132kB)

Abstract

In the current world of marketing, the importance of using the data that companies collect to their advantage is rising. Data mining makes the process of analysing the data and providing insights more accessible and much more efficient. We will adapt existing models, a sailfish optimization algorithm with random dis- turbance strategy - extreme learning machine (SFOR-ELM) and Long Short-Term Mem- ory network (LSTM), for data prediction to our cause and data, and we will try to predict the sales per month of products the company offers, as well as the total amount of sales a week. These predictions are going to help the company with planning and management processes, and thus save them money as well as time. Later we will compare these models for the prediction based on the accuracy and analyze which of them is more suitable for the data we were given. Secondly, we will use statistical tools, to provide customer segmentation of the company customers, providing an overview of the customers, helping the company to understand them better, and selecting groups of customers that the marketing should be focused on. The paper demonstrates how to perform customer profiling using the RFM model and predict customers’ churn with almost 96% accuracy. Furthermore, the predictions of the total amount of orders in a week with mean squared error of 0,156 and mean average error of 0,307 and the predictions of parts sold in a month with mean squared error of 0,15 and mean average error of 0,11.

Item Type:	Thesis (Bachelor's Thesis)
Supervisor name:	Dustegor, D. and Truong, H.C.
Degree programme:	Computing Science
Thesis type:	Bachelor's Thesis
Language:	English
Date Deposited:	09 Aug 2024 07:29
Last Modified:	09 Aug 2024 07:29
URI:	https://fse.studenttheses.ub.rug.nl/id/eprint/33916

Actions (login required)

View Item