Sosyal Bilimler Enstitüsü

Browse

Now showing 1 - 1 of 1

Analysis of online marketplace sales prediction based on machine learning algorithms: A case of Turkish e-commerce site
(Abdullah Gül Üniversitesi / Sosyal Bilimler Enstitüsü, 2023) Kaya, Ecem; AGÜ, Sosyal Bilimler Enstitüsü, İşletme ve Ekonomi İçin Veri Bilimi Ana Bilim Dalı
Internet shopping has grown in popularity as more of our daily requirements have begun to be addressed online. Learning about the preferences and motivations of customers in the Turkish market and guiding e-commerce platforms to adapt their marketing strategies and increase customer satisfaction is important for both resource allocation and cost minimization. The purpose of this paper is to estimate future sales for popular e-commerce sites based on behavioral factors such as discounts, price or free shipping. Therefore, real-time and experiment-independent data are collected from the sales made by one of Turkey's most popular e-commerce sites. In order to produce predictions, we employ Artificial Neural Networks, Support Vector Regression, K-Nearest Neighbors Regressor, OLS regression, and Nu-Support Vector Regressor. The models developed using machine learning algorithms attempt to estimate the number of sales based on independent factors such as price, discount rate, and user ratings. As the result of this research, we calculate and compare the accuracy of the models with root mean squared errors and R².