Suicidal Ideation Detection (NLP)
Problem: Early identification of suicidal ideation from large-scale social text data.
Stack: Python, Scikit-learn, TF-IDF, NLP
92.9% accuracy and 98.2% ROC-AUC on 232K Reddit posts.
Supports mental-health initiatives through early warning insights.
Read full case study