A novel three-layer stacked Machine Learning system to detect financial opportunities, i.e. positive user speculations and forecasts about stock market assets, in micro-blogging data with high precision.
A representation learning framework that captures diverse local and global properties of financial transactional data, and incorporates external context to further improve the quality of the representations.
A novel method for missing data imputation using granular semantics and an AI-driven pipeline to accurately predict bankruptcy in large, high-dimensional, and imbalanced financial datasets.