Related Experiment Video
Updated: Mar 1, 2026

Project-Based Learning Guidelines for Health Sciences Students: An Analysis with Data Mining and Qualitative Techniques
Published on: December 9, 2022
LQG Online Learning
Giorgio Gnecco1, Alberto Bemporad2, Marco Gori3
1DYSCO Research Unit, IMT School for Advanced Studies, Piazza S. Francesco, 19-55110 Lucca, Italy giorgio.gnecco@imtlucca.it.
This study introduces a novel online learning framework using optimal control and machine learning. The method offers smoother parameter estimates than the Kalman filter, showing improved robustness to outliers.
Area of Science:
- Control Theory
- Machine Learning
- Statistical Signal Processing
Background:
- Online learning from supervised examples is crucial for adaptive systems.
- Classical methods like the Linear Quadratic Gaussian (LQG) problem provide a foundation.
- Existing methods can be sensitive to outliers and lack robustness.
Purpose of the Study:
- To develop a novel optimal control formulation for online supervised learning.
- To investigate the relationship between this new framework and LQG control.
- To enhance robustness and smoothness of parameter estimates in online learning.
Main Methods:
- Combining optimal control theory with machine learning techniques.
- Formulating and solving an optimal control problem for online learning with regularization.
- Comparing the proposed method with the Kalman filter for parameter estimation.
Main Results:
- The proposed algorithm provides closed-form optimal solutions for online learning.
- The method demonstrates greater robustness to outliers compared to the Kalman filter due to regularization.
- Smoother time-varying parameter estimates are achieved.
Conclusions:
- The developed optimal control framework offers a robust and efficient approach to online supervised learning.
- Regularization is key to improving outlier resilience and estimate smoothness.
- Extensions to infinite horizons and nonlinear models are feasible.
Related Concept Videos
Purposive Learning
Cognitive Learning
E. C. Tolman's theory of purposive behavior emphasizes that much behavior is goal-directed. He argued that to understand behavior, we must look at the entire sequence of actions leading to a goal. For instance, high school students study hard, not just due to past reinforcement but also to achieve the goal of getting into a good college.
Tolman introduced the idea that behavior is influenced by...
Introduction to Learning
In contrast to learned behaviors, unlearned behaviors such as crying, sexual...
Learning Disabilities
Dyslexia
Dyslexia is a...
Observational Learning
Quality Assurance

