Intelligent Systems Lab Amsterdam

Learning

Intelligent systems need to learn in order to cope with changing environments or inaccurate a-priori knowledge. Learning from data is studied in the context of autonomous systems. Originally we focussed on neural networks, now we look more general to probabilistic methods.

• COGNIRON: a Cognitive Robot Companion
• Planning under uncertainty
• Distributed surveillance systems and probabilistic models
• Audio visual data fusion

• Aircraft recognition from radar profiles
• Environment learning and map building
• Nonlinear projection methods
• Personal robot in intelligent environment
• Learning Concepts in Real World Embodied Agents
• Distributed User Modeling for Personalised Exploring Recommender Systems