Peak Performance in Fake News Detection: Unveiling the Potential of Deep Learning Optimizers

The World Wide Web's introduction and the quick uptake of social media sites like Facebook and Twitter opened the door to a level of information sharing never seen in human history. Social media platforms are being used by consumers to create and share more information than ever before, some of it false and unrelated to reality. It is difficult to automatically classify a text article as misleading or disinformation. Even a subject-matter expert must consider a variety of factors before determining if an article is true. In this work, we propose to automatically classify news articles using an ensemble machine learning approach. Our research examines many textual characteristics that can be utilized to discern authentic content from counterfeit. We train a variety of machine learning algorithms employing those properties in conjunction with different ensemble approaches, and we assess their performance on four real-world datasets. Our suggested ensemble learner technique outperforms individual learners, as confirmed by experimental evaluation.