The vast quantity of information brought by big data as well as the evolving computer hardware encourages success stories in the machine learning community. In the meanwhile, it poses challenges for the Gaussian process (GP) regression, a well-known non-parametric and interpretable Bayesian model, which suffers from cubic complexity to data size. To improve the scalability while retaining desirable prediction quality, a variety of scalable GPs have been presented. But they have not yet been comprehensively reviewed and analyzed in order to be well understood by both academia and industry. The review of scalable GPs in the GP community is timely and important due to the explosion of data size. To this end, this paper is devoted to the review on state-of-the-art scalable GPs involving two main categories: global approximations which distillate the entire data and local approximations which divide the data for subspace learning. Particularly, for global approximations, we mainly focus on sparse approximations comprising prior approximations which modify the prior but perform exact inference, posterior approximations which retain exact prior but perform approximate inference, and structured sparse approximations which exploit specific structures in kernel matrix; for local approximations, we highlight the mixture/product of experts that conducts model averaging from multiple local experts to boost predictions. To present a complete review, recent advances for improving the scalability and capability of scalable GPs are reviewed. Finally, the extensions and open issues regarding the implementation of scalable GPs in various scenarios are reviewed and discussed to inspire novel ideas for future research avenues.Index Terms-Gaussian process regression, big data, scalability, sparse approximations, local approximations Haitao Liu is with the Rolls-

No abstract

Based on the characteristics of electric vehicles (EVs), this paper establishes the load models of EVs under the autonomous charging mode and the coordinated charging and discharging mode. Integrating the EVs into a microgrid system which includes wind turbines (WTs), photovoltaic arrays (PVs), diesel engines (DEs), fuel cells (FCs) and a storage battery (BS), this paper establishes multi-objective economic dispatch models of a microgrid, including the lowest operating cost, the least carbon dioxide emissions, and the lowest pollutant treatment cost. After converting the multi-objective functions to a single objective function by using the judgment matrix method, we analyze the dynamic economic dispatch of the microgrid system including vehicle-to-grid (V2G) with an improved particle swarm optimization algorithm under different operation control strategies. With the example system, the proposed models and strategies are verified and analyzed. Simulation results show that the microgrid system with EVs under the coordinated charging and discharging mode has better operation economics than the autonomous charging mode. Meanwhile, the greater the load fluctuation is, the higher the operating cost of the microgrid system is.

No abstract

No abstract

No abstract

Stephania tetrandra S. Moore (S. tetrandra) is distributed widely in tropical and subtropical regions of Asia and Africa. The root of this plant is known in Chinese as ''Fen Fang Ji''. It is commonly used in traditional Chinese medicine to treat arthralgia caused by rheumatism, wet beriberi, dysuria, eczema and inflamed sores. Although promising reports have been published on the various chemical constituents and activities of S. tetrandra, no review comprehensively summarizes its traditional uses, phytochemistry, pharmacology and toxicology. Therefore, the review aims to provide a critical and comprehensive evaluation of the traditional use, phytochemistry, pharmacological properties, pharmacokinetics and toxicology of S. tetrandra in China, and meaningful guidelines for future investigations.

Summary In this paper, a solution to the continuous output‐feedback finite‐time control problem is proposed for a class of second‐order MIMO nonlinear systems with disturbances. First, a continuous finite‐time controller is designed to stabilize system states at equilibrium points in finite time, which is proven correct by a constructive Lyapunov function. Next, because only the measured output is available for feedback, a continuous nonlinear observer is presented to reconstruct the total states in finite time and estimate the unknown disturbances. Then, a continuous output‐feedback finite‐time controller is proposed to track the desired trajectory accurately or alternatively converge to an arbitrarily small region in finite time. Finally, proposed methods are applied to robotic manipulators, and simulations are given to illustrate the applicability of the proposed control approach. Copyright © 2015 John Wiley & Sons, Ltd.