HateSpeech-DetectAI

Built a hate speech detection system using NLP and ML models like SimpleRNN, LSTM, BERT, XGBoost, and ensemble classifiers (Bagging, Boosting, Stacking, Voting). BERT achieved top F1 score (0.81), with ensemble methods adding stability. Conducted EDA and model evaluation (AUC, F1, ROC) to ensure accuracy, creating a robust solution.