Applying Machine Learning to Predict Online Customers Behaviour

The way of learning continues to develop and gradually changes the way people buy such as online learning sites. This has implications for the strategies used for marketing in order to increase users on their platform. As these changes change, predicting consumer behavior and choices is becoming a topic of interest to researchers and companies alike. Predicting consumer behavior patterns has been conventionally proven, especially among sales, to increase business growth and generate customer loyalty. The study analyzed customer behavior patterns on paid online learning sites using Google Data Analytics, while in February 2022, a digital report by Hootsuite recorded at least 44% of Internet usage in Indonesia for E-learning needs. The use of Machine Learning is increasingly glimpsed to provide convenience in various aspects of work. We use preprocessing oversampling and feature selection steps to improve classifier performance and scalability. In the case of this paper, the authors used Decision Tree, Random Forest, XG Boosting, KNN, and SVM algorithms through a different set of conditions and then proposed an understandable and highly accurate machine learning model. This paper intends to analyze consumer behavior patterns and understand the metrics that determine consumers in making decisions using sensitive analysis. We also provide marketing strategies and further propose the use of machine learning algorithms in predicting customer behavior patterns. Further results can be extended to various departments to evaluate performance on the platform to increase customer loyalty