<![CDATA[On an online shopping website, the platform may provide a service to the shop owners by suggesting which items to promote. One possible consideration is price. If an item is priced more expensively than the average price of other items in the same category, then the item should be advertised more intensely, or repriced. Due to the quickly growing number of products and categories, calculating the average price in real time can be difficult or slow. Alternatively, one may employ machine learning algorithms. In this study, we use Extra Trees Classifier on clickstream data, which is user activity report. We demonstrate the algorithm on the clickstream data of a an online shopping website for pregnant women, retrieved from UCI Machine Learning Repository Dataset. The data has 14 attributes and 165474 entries. The model is trained on 75% of the data, and tested on the remaining 25%, with an observed accuracy of 99 %.]]>
Copyrights © 2022
