This study examines sentiment analysis methods to interpret audience reactions to educational videos on YouTube. It analyzes approximately 104,825 comments using the VADER (Valence Aware Dictionary and Sentiment Reasoner) lexicon-based model alongside various machine learning classifiers to ascertain the polarity of sentiment. Algorithms such as the Passive Aggressive Classifier, Logistic Regression, Support Vector Machines (SVM), Multinomial Naïve Bayes (MultinomialNB), and Random Forest are employed for classification, while model effectiveness is assessed using confusion matrix analysis. Furthermore, frequently used terms in positive, neutral, and negative sentiments are identified to explore shared viewer expressions across different sentiment categories. Word clouds offer visual insight into commonly used terms, enhancing our understanding of the audience's feelings toward educational content. The findings provide extensive insight into audience perceptions and responses, highlighting the potential of using sentiment analysis to enhance educational material on YouTube. The study provides important insights for YouTube creators, educational policymakers, developers of social networking sites, and analysts in understanding audience feelings and effectively designing content strategies.