Sentiment analysis is a crucial natural language processing (NLP) task that automates the assessment of emotions in text. It has wide applications in social media, reviews, and customer feedback. This paper presents a comprehensive comparative study on sentiment analysis techniques, encompassing multinomial Naive Bayes, SVM, random forest, gradient boosting, voting classifier, LSTM and BERT. The evaluation includes precision, accuracy, classification reports, as well as feature extraction and lemmatization. This analysis covers both traditional machine learning and deep learning methods, providing a refined understanding of their performance for practical sentiment analysis applications.