A Hybrid Deep Feature Extraction and Gradient Boosting Framework for Robust Fake News Detection

A survey in which classical machine learning, deep learning, ensemble, and hybrid models are compared for fake news detection . Alternative hybrid framework that combines deep semantic features with classification based on tree ensembles, specifically XGBoost . Evaluation on diverse datasets focusing on the importance of data quality and the distribution in .Error and robustness analysis providing insight into overfitting, performance in challenging scenarios.