Using Deep Learning, NLP

This project focuses on predicting machine failures using sensor data collected from industrial environments. The dataset includes various real-time indicators such as air quality, temperature, current usage, pressure, and movement data, which reflect the operating condition of machines.

TriFuse-PdM is an interpretable, fusion-based predictive maintenance pipeline combining classical ML, deep learning, and SHAP explainability with advanced resampling and calibration for robust industrial failure prediction.