Search

Remaining useful life prediction is an important way to improve the durability and reduce the cost of the proton exchange membrane fuel cell. This paper presents a novel method to predict the remaining useful life of proton exchange membrane fuel cell under different load currents based on support vector regression and grey wolf optimizer algorithm. The proposed method considers the influence of 17 operating conditions and historical voltage. Firstly, the measured data are reconstructed through robust locally weighted smoothing method to reduce the calculation amount and filter disturbances. Then, support vector regression with fewer hyperparameters is used to establish the degradation model. Finally, the hyperparameters of support vector regression are optimized through grey wolf optimizer algorithm to improve the accuracy of degradation prediction. The proposed method is validated by two degradation experiments under different load currents. The test results show that grey wolf optimizer algorithm can effectively improve the accuracy of degradation prediction based on support vector regression. Compared with other methods, the proposed method has the highest accuracy. The proposed method can predict the fuel cell degradation with a mean absolute percentage error of less than 0.3%. The proposed method can predict the remaining useful life of 492 hours. [ABSTRACT FROM AUTHOR]

In this paper, we present a framework for performing automatic analysis of Land Use Zones based on Location-Based Social Networks (LBSNs). We model city areas using a hierarchical structure of POIs extracted from foursquare. We encode such structures in kernel machines, e.g., Support Vector Machines, using a new Tree Kernel, i.e., the Hierarchical POI Kernel (HPK), which can take the importance of the individual POIs into account during the substructure matching. This way, HPK projects structures in the space of all their possible substructures such that each dimension corresponds to a semantic structural feature, weighted according to the discriminative power of POIs. We generated four different datasets for the following cities: Barcelona, Lisbon, Amsterdam and Milan, where we trained and tested our models. The results show that our approach largely outperforms previous work and standard baseline built on simple features, such as counts of different POIs. Finally, we apply a mining algorithm to extract the most relevant features (tree fragments) from the implicit TK space according to the weights the kernel machine assigned to them. Our approach can produce an explicit set of representative features that can be used to classify and characterize urban areas. [ABSTRACT FROM AUTHOR]

Multi-label classification is an important research topic in machine learning, for which exploiting label dependencies is an effective modeling principle. Recently, probabilistic models have shown great potential in discovering dependencies among labels. In this paper, motivated by the recent success of multi-view learning to improve the generalization performance, we propose a novel multi-view probabilistic model named latent conditional Bernoulli mixture (LCBM) for multi-label classification. LCBM is a generative model taking features from different views as inputs, and conditional on the latent subspace shared by the views a Bernoulli mixture model is adopted to build label dependencies. Inside each component of the mixture, the labels have a weak correlation which facilitates computational convenience. The mean field variational inference framework is used to carry out approximate posterior inference in the probabilistic model, where we propose a Gaussian mixture variational autoencoder (GMVAE) for effective posterior approximation. We further develop a scalable stochastic training algorithm for efficiently optimizing the model parameters and variational parameters, and derive an efficient prediction procedure based on greedy search. Experimental results on multiple benchmark datasets show that our approach outperforms other state-of-the-art methods under various metrics. [ABSTRACT FROM AUTHOR]

During the running of electric locomotives, a large amount of operating data will be stored through the Analog-Digital Converter (ADC) and the recording oscillograph of the on-board Traction Control Unit (TCU). Analysis based on mass data to detect slipping is of great significance for saving energy, improving the performance and safety of the traction control system of electric locomotives. Therefore, after analyzing and summarizing the advantages of Empirical Wavelet Transform (EWT) and Fuzzy Entropy (FE) algorithm in data processing and the classification theory of Support Vector Machine (SVM), this paper proposed a SVM slipping detection method based on EWT and FE algorithm. Firstly, feature extraction is carried out on the locomotive wheelset velocity data collected by the on-board velocity sensor, that is, the feature vector of data is extracted by calculating fuzzy entropy value after EWT, and the velocity data is converted into feature matrix for slipping-detection of SVM. Then considering the defects of traditional optimization algorithms such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) in the parameter optimization of SVM, Genetic Particle Swarm Optimization Algorithm (GAPSO) is proposed to optimize the parameters of SVM classification model, and the slipping detection model of SVM is established through training samples, which realizes high-precision slipping detection. Finally, combined with a large number of locomotive data, the comparative analysis test is carried out, the results show that the feature extraction method based on EWT and FE algorithm has excellent ability of mode decomposition and feature extraction, GAPSO has efficient optimization ability, and the proposed slipping detection method can accurately detect slipping. [ABSTRACT FROM AUTHOR]

As the training dataset size and the model size of machine learning increase rapidly, more computing resources are consumed to speedup the training process. However, the scalability and performance reproducibility of parallel machine learning training, which mainly uses stochastic optimization algorithms, are limited. In this paper, we demonstrate that the sample difference in the dataset plays a prominent role in the scalability of parallel machine learning algorithms. We propose to use statistical properties of dataset to measure sample differences. These properties include the variance of sample features, sample sparsity, sample diversity, and similarity in sampling sequences. We choose four types of parallel training algorithms as our research objects: (1) the asynchronous parallel SGD algorithm (Hogwild! algorithm), (2) the parallel model average SGD algorithm (minibatch SGD algorithm), (3) the decentralization optimization algorithm, and (4) the dual coordinate optimization (DADM algorithm). Our results show that the statistical properties of training datasets determine the scalability upper bound of these parallel training algorithms. [ABSTRACT FROM AUTHOR]