Using machine learning to predict sellers’ activity next month at an online auction platform

No Thumbnail Available
Journal Title
Journal ISSN
Volume Title
School of Business | Master's thesis
Date
2022
Major/Subject
Mcode
Degree programme
Information and Service Management (ISM)
Language
en
Pages
40 + 4
Series
Abstract
The survival and growth of online auction platform depends on the ability of predicting the sellers’ activity as accurately as possible. With the results of determining how sellers will be active in the future, the marketing effort and many operational decisions will target the right sellers at the right time. This thesis attempts to identify how to measure the sellers’ activity and how it will be next month at an online auction platform by using machine learning models. The model is built on the historical data on the platform, which is collected and aggregated by SQL. After identifying what sellers’ activity is and how to measure it, Linear Regression and Random Forest Regressor are applied to the data. The performance of two algorithms will be compared to achieve the model with the highest accuracy. The result is the reusable model that predicts the sellers’ activity next month and the recommendations how to improve the model in the next research.
Description
Thesis advisor
Liu, Yong
Yanqing, Lin
Keywords
machine learning, online auction, sellers' activity, random forest regressor, linear regression, activity prediction, feature engineering
Other note
Citation