Summary Machine learning has been heavily researched and widely used in many disciplines. However, achieving high accuracy requires a large amount of data that is sometimes difficult, expensive, or impractical to obtain. Integrating human knowledge into machine learning can significantly reduce data requirement, increase reliability and robustness of machine learning, and build explainable machine learning systems. This allows leveraging the vast amount of human knowledge and capability of machine learning to achieve functions and performance not available before and will facilitate the interaction between human beings and machine learning systems, making machine learning decisions understandable to humans. This paper gives an overview of the knowledge and its representations that can be integrated into machine learning and the methodology. We cover the fundamentals, current status, and recent progress of the methods, with a focus on popular and new topics. The perspectives on future directions are also discussed.

No abstract

The purpose of this paper is to introduce an experimental method for investigating the motion of an electric unicycle and its rider. As the balancing and maneuvering method of human on electric unicycle is known, this may help us in evaluating the safety of human rider, design balance-assist features and/or self-driving functionalities.

No abstract

An on-board traffic prediction algorithm is proposed for connected vehicles traveling on highways. The prediction is based on data received from other connected vehicles ahead in the traffic stream, leveraging the fact that a vehicle will enter the traffic that other vehicles ahead have already met. Our method includes traffic state estimation with Kalman filter and prediction via traffic flow models describing the propagation of congestion waves. The end result is an individualized speed preview in real time up to about half a minute for the connected vehicle executing prediction. Most importantly, the traffic prediction was successfully implemented on board of a real vehicle and predictions were tested in real traffic with experiments involving connected human-driven vehicles.