Fake News Classification

This project explores binary classification of news articles as real or fake using two machine learning models: Naive Bayes and Support Vector Classifier (SVC). Exploratory data analysis included visualizing class distributions, generating word clouds, and examining feature importance to gain insight into the data. Model evaluation was conducted using a confusion matrix, precision-recall curves, and a full classification report.

Key Results:

The Support Vector Classifier outperformed Naive Bayes across precision, recall, and F1-score metrics, indicating stronger predictive performance.

Feature analysis identified terms and patterns commonly associated with misinformation, improving model interpretability and understanding of the dataset.