Comparing Machine Learning Classification Algorithms and Feed Forward Neural Network to Perform Sentiment Analysis

We have utilized a unique dataset of 50,000 review inputs in this research to perform sentiment analysis using a selection of machine learning classification methods. To eliminate potential unnecessary terms in the reviews, the dataset is extensively preprocessed prior to getting employed by any algorithm. In order to get the best results, we evaluated numerous classification techniques, including Logistic Regression, Random Forest, Extra Tree Classifier, Support Vector Machine (SVM), and Naïve Bayes. The outputs of all the underlying algorithms are then aggregated by feeding them through a Voting Classification model. Furthermore, to make a clear contrast, we designed a Feed Forward Neural Network Engine as well. The data in the Neural Network is routed via various layers, where it is refined to get the best outcomes.