Machine learning is a branch of artificial intelligence (AI) and computer science which focuses on the use of data and algorithms to imitate the way that humans learn and gradually improve accuracy. Interestingly several tools for machine learning are very closely connected to adaptive control and system identification (e.g. parameter estimation, recursive estimation, online minimization of a performance error). Building on these connections and common concepts, new intersections and opportunities for improved machine learning algorithms naturally arise.
Representative themes
- Deep learning
- Extreme learning and broad learning
- Recursive learning
- Lifelong learning and continuous learning
Representative applications
- Computer vision, image processing
- Automated driving
- Traffic prediction, Energy consumption prediction
- Prediction of human behavior (for human-in-the-loop applications)
- Simultaneous localization and mapping (SLAM)
Representative results
- Liu D., Baldi S., Yu W., and Chen C. L. P., “A hybrid recursive implementation of broad learning with incremental features”, IEEE Transactions on Neural Networks and Learning Systems, 2021 (co-first author) doi:10.1109/TNNLS.2020.3043110
- Liu D., Baldi S., Yu W., Cao J., and Huang W., “On training traffic predictors via broad learning structures: a benchmark study”, IEEE Transactions on Systems, Man, and Cybernetics: Systems, Vol. 52(2), pp. 749-758, 2022. doi:10.1109/TSMC.2020.3006124
- Moerland T. M., Deichler A., Baldi S., Broekens J., and Jonker C. M., “Think neither too fast nor too slow: the computational trade-off between planning and reinforcement learning”, The International Conference on Automated Planning and Scheduling (ICAPS 2020), October 26th-30th, Nancy, France, 2020.