AI-powered mental health monitoring system using NLP, MentalBERT, PCA, and XGBoost for early risk assessment.
-
Updated
Jun 18, 2026 - JavaScript
AI-powered mental health monitoring system using NLP, MentalBERT, PCA, and XGBoost for early risk assessment.
A collection of NLP projects built with HuggingFace Transformers, TensorFlow/Keras, and classical methods. Each notebook covers a distinct task and architecture.
Research exploring how hyperparameter tuning, class-specific data augmentation, and MentalBERT's domain-specific architecture each contribute to improving multi-class mental health text classification and minority class detection.
Clinical-grade NLP triage system with Nested-CV-validated ML and grounded LLM copilot. Streamlit dashboard.
To associate your repository with the mentalbert topic, visit your repo's landing page and select "manage topics."