transparent ML





Machine learning for a better world

Machine learning is out to the world. Much of what previously programmed manually can now be learned through examples. 


  • Explainable AI: Machine learning is often criticised as black-box which is perhaps the most critical point that blocks wide adoption. This research aims to improve the situation by creating a transparent-box.
  • Recomemender systems: The goal of a recommender system is to deliver right services to right users. Most existing work is rather ad-hoc and ignores complex nature of the data. Research topics include modeling choices, discovering hidden patterns, incorporating contexts and side-information, social networks, multiple-domains, product hierarchies, as well as correlations between actors and items.
  • Automated software engineering: Software is eating the world. Automated software engineering helps improving software development with better risk estimation, lower cost and higher code quality.
  • Anomaly detection: Here we learn to detect unknown unknowns (e.g., outliers and anomalies).
  • Process modeling: Processes are logics that underpin businesses, factories and operations. ML offers new ways for discovery of actionable insights from event log, as well as process prediction, optimization and planning.


Explainable AI

Recommender systems
Automated software engineering
Anomaly detection
Process modeling