A Forecast of Consumer Behaviour using the Naivesbayes Algorithm

Predicting customer behavior is essential for developing an effective e-commerce strategy so that companies can offer personal and targeted marketing based on likely purchase trends. This paper proposes a model based on the Naive Bayes classifier to predict purchases based on customer purchase behavior defined by the following attributes: age, gender, predicted salary, and purchase history. The Naive Bayes algorithm has been selected due to its simplicity and efficiency to handle the task of multifeature classification. The data preprocessing included categorical encoding and normalization of the continuous ones. Then, this dataset was divided into a training and testing subset. For this model, prediction about the likelihood to buy had already reached an accuracy of 81% More metrics used for evaluation were precision, recall, the F1-score: all gave a thumbs up at its capability to handle class imbalance appropriately. The study findings indicated that income and age were the two major factors that influenced a customer's purchasing behavior, though the higher-income customers and greater age groups tend to make more purchases. These research findings thus prove that the Naive Bayes classifier has practical value in real-life ecommerce applications for the tasks of customer segmentation, thereby supporting developments in more sophisticated marketing strategies.