An Experimental Study on Prediction of Video Ads Engagement

In the realm of digital advertising, the quest to anticipate and adapt to user engagement patterns with video advertisements drives the adoption of advanced machine learning techniques. This study utilizes a rich dataset encompassing various user attributes such as demographic information, browsing behavior, and device details. Leveraging Principal Component Analysis (PCA) for feature extraction, the dataset's dimensionality is reduced while preserving crucial patterns, thus enhancing the effectiveness of the predictive model. Through rigorous training and evaluation, employing metrics like accuracy, confusion matrix analysis, precision, recall, and F1 score, the study showcases the efficacy of the employed machine learning approach, particularly in predicting user engagement with video ads. These findings underscore the potential of machine learning in providing actionable insights into user behavior, guiding advertising strategies and decision-making processes. Furthermore, the application of PCA in feature extraction stands out as a notable contribution, refining the dataset and improving prediction accuracy. By delving into user engagement analysis, this research offers valuable insights for advertisers aiming to thrive in the competitive landscape of digital advertising.