The Experts below are selected from a list of 30978 Experts worldwide ranked by ideXlab platform
Xin Chen - One of the best experts on this subject based on the ideXlab platform.
-
Gaussian Kernel based adaptive critic design using two phase value iteration
Information Sciences, 2019Co-Authors: Xin Chen, Wei Wang, Weihua CaoAbstract:Abstract Adaptive critic design is an efficient way to learn optimal action policies on-line, in which a critic network plays an important role to estimate value functions. Because of its good generalization and easy configuration, Kernel-based method is prevalently introduced to the construction of critic network. Conventionally the hyper-parameters of Kernel-based model need to be predetermined, but empirical selection of them may mislead Kernel-based regression with an improper modeling hypothesis space. To tackle this problem, a two-phase iteration of value function approximation and hyper-parameters optimization for Gaussian-Kernel based adaptive critic design (GK-ACD) is presented in this paper, which not only approximates the value functions, but also updates the hyper-parameters on-line. Since the two phases are strong coupling, the theoretical proof based on stochastic approximation derives the sufficient conditions guaranteeing the convergence, and points out that the algorithm’s performance mostly relies on the design of coordinated learning rates w.r.t. the two phases. Finally a series of numerical experiments are given to discuss the necessity of two-phase updates and the performance under the coordinated learning rates.
-
two phase iteration for value function approximation and hyperparameter optimization in Gaussian Kernel based adaptive critic design
Mathematical Problems in Engineering, 2015Co-Authors: Xin Chen, Penghuan Xie, Yonghua XiongAbstract:Adaptive Dynamic Programming (ADP) with critic-actor architecture is an effective way to perform online learning control. To avoid the subjectivity in the design of a neural network that serves as a critic network, Kernel-based adaptive critic design (ACD) was developed recently. There are two essential issues for a static Kernel-based model: how to determine proper hyperparameters in advance and how to select right samples to describe the value function. They all rely on the assessment of sample values. Based on the theoretical analysis, this paper presents a two-phase simultaneous learning method for a Gaussian-Kernel-based critic network. It is able to estimate the values of samples without infinitively revisiting them. And the hyperparameters of the Kernel model are optimized simultaneously. Based on the estimated sample values, the sample set can be refined by adding alternatives or deleting redundances. Combining this critic design with actor network, we present a Gaussian-Kernel-based Adaptive Dynamic Programming (GK-ADP) approach. Simulations are used to verify its feasibility, particularly the necessity of two-phase learning, the convergence characteristics, and the improvement of the system performance by using a varying sample set.
-
online clutter estimation using a Gaussian Kernel density estimator for multitarget tracking
Iet Radar Sonar and Navigation, 2015Co-Authors: Xin Chen, R Tharmarasa, Thia Kirubarajan, M McdonaldAbstract:In this study, the spatial distribution of false alarms is assumed to be a non-homogeneous Poisson point (NHPP) process. Then, a new method is developed under the Kernel density estimation (KDE) framework to estimate the spatial intensity of false alarms for the multitarget tracking problem. In the proposed method, the false alarm spatial intensity estimation problem is decomposed into two subproblems: (i) estimating the number of false alarms in one scan and (ii) estimating the variation of the intensity function value in the measurement space. Under the NHPP assumption, the only parameter that needs to be estimated for the first subproblem is the mean of false alarm number, and the empirical mean is used here as the maximum likelihood estimate of that parameter. Then, for the second subproblem, an online multivariate local adaptive Gaussian Kernel density estimator is proposed. Furthermore, the proposed estimation method is seamlessly integrated with widely used multitarget trackers, like the joint integrated probabilistic data association algorithm and the multiple hypotheses tracking algorithm. Simulation results show that the proposed KDE-based method can provide a better estimate of the false alarm spatial intensity and help the multitarget trackers yield superior performance in scenarios with spatially non-homogeneous false alarms.
-
online clutter estimation using a Gaussian Kernel density estimator for target tracking
International Conference on Information Fusion, 2011Co-Authors: Xin Chen, R Tharmarasa, Thia Kirubarajan, Michel PelletierAbstract:In this paper, based on non-homogeneous Poisson point processes (NHPP), a Kernel clutter spatial intensity estimation method is proposed. Here, the clutter spatial intensity estimation problem is decomposed into two parts: (1) estimate the probability distribution of the clutter number per scan; (2) estimate the spatial variation of the clutter intensity in the measurement space. Under the NHPP assumption, the empirical mean is used to get a maximum likelihood estimate for the first problem. For the second problem, an online locally adaptive Gaussian Kernel density estimator is proposed. In addition, the proposed clutter estimation method is integrated with standard multitarget trackers, like Multiple Hypothesis Tracker (MHT), Joint Integrated Probabilistic Data Association (JIPDA) tracker, Probability Hypothesis Density (PHD) filter. Simulation results show that the proposed clutter spatial intensity estimator can improve the performance of the multitarget tracker in the presence of non-homogeneous clutter background.
Weihua Cao - One of the best experts on this subject based on the ideXlab platform.
-
Gaussian Kernel based adaptive critic design using two phase value iteration
Information Sciences, 2019Co-Authors: Xin Chen, Wei Wang, Weihua CaoAbstract:Abstract Adaptive critic design is an efficient way to learn optimal action policies on-line, in which a critic network plays an important role to estimate value functions. Because of its good generalization and easy configuration, Kernel-based method is prevalently introduced to the construction of critic network. Conventionally the hyper-parameters of Kernel-based model need to be predetermined, but empirical selection of them may mislead Kernel-based regression with an improper modeling hypothesis space. To tackle this problem, a two-phase iteration of value function approximation and hyper-parameters optimization for Gaussian-Kernel based adaptive critic design (GK-ACD) is presented in this paper, which not only approximates the value functions, but also updates the hyper-parameters on-line. Since the two phases are strong coupling, the theoretical proof based on stochastic approximation derives the sufficient conditions guaranteeing the convergence, and points out that the algorithm’s performance mostly relies on the design of coordinated learning rates w.r.t. the two phases. Finally a series of numerical experiments are given to discuss the necessity of two-phase updates and the performance under the coordinated learning rates.
Peng Zou - One of the best experts on this subject based on the ideXlab platform.
-
nonlinear predistortion scheme based on Gaussian Kernel aided deep neural networks channel estimator for visible light communication system
Optical Engineering, 2019Co-Authors: Yiheng Zhao, Meng Shi, Peng Zou, Nan ChiAbstract:A scheme for Gaussian Kernel-aided deep neural networks nonlinear predistortion (GK-DNNPD), which could effectively reduce the computational complexity of the receivers, is experimentally demonstrated. Compared with lookup table (LUT) PD, the GK-DNNPD could increase the Q -factor of 8 pulse amplitude modulation visible light communication (VLC) system by 1.56 dB at 1.335 Gbps. We experimentally proved that GK-DNNPD could increase the bitrate under hard-decision forward error correction from 1.335 to 1.385 Gbps. This is the first time that GK-DNNs are utilized for PD in the field of VLC systems. Meanwhile, GK-DNNPD requires less data for training than LUT, and the space complexity of the model is lower than LUT as well, which provides GK-DNNPD with the potential to be applied in practical VLC systems.
-
Gaussian Kernel aided deep neural network equalizer utilized in underwater pam8 visible light communication system
Optics Express, 2018Co-Authors: Nan Chi, Yiheng Zhao, Meng Shi, Peng ZouAbstract:In this paper, we demonstrate a novel Gaussian Kernel-aided deep neural network (GK-DNN) equalizer that can effectively compensate for the high nonlinear distortion of underwater PAM8 visible light communication (VLC) channels. The application of a Gaussian Kernel can reduce the necessary training iterations to 47.06%, enabling it to outperform the traditional DNN equalizer. At the same time, a novel design strategy with respect to the structure of the GK-DNN equalizer is proposed, which can effectively save computing resources and reduce the data volume of the necessary training data set. By using the GK-DNN equalizer, a 1.5 Gbps PAM8 VLC system over 1.2-m underwater transmission is successfully demonstrated.
Tadashi Shibata - One of the best experts on this subject based on the ideXlab platform.
-
an on chip trainable Gaussian Kernel analog support vector machine
IEEE Transactions on Circuits and Systems, 2010Co-Authors: Kyunghee Kang, Tadashi ShibataAbstract:An analog circuit architecture of Gaussian-Kernel support vector machines having on-chip training capability has been developed. It has a scalable array processor configuration and the circuit size increases only in proportion to the number of learning samples. Thanks to the hardware-friendly algorithm employed in the present work, the learning function is realized by attaching a small additional circuitry to the SVM classifying hardware. The SVM classifying hardware is composed as an array of Gaussian circuits. Although the system is inherently analog, the input and output signals including training results are all available in digital format. Therefore, the learned parameters are easily stored and reused after training sessions. A proof-of concept chip containing 2-class, 2-D, 12-template classifier was designed and fabricated in a 0.18-μm CMOS technology. The experimental results obtained from the fabricated chips are presented and compared with theoretical calculation results. It can classify 8.7 x 105 vectors per second and the average power dissipation was 220 μW. The learning capability was tested using eight fabricated chips and the variability among these chips were evaluated. Successful operation of the chips was confirmed by measurement results, which demonstrates that on-chip-learning can compensate for analog imperfections.
-
an on chip trainable Gaussian Kernel analog support vector machine
International Symposium on Circuits and Systems, 2009Co-Authors: Kyunghee Kang, Tadashi ShibataAbstract:We propose an analog circuit architecture of the Gaussian-Kernel support vector machine having on-chip training capability. Thanks to the hardware-friendly algorithm, the learning function is realized by attaching a small additional circuitry to the SVM classifying hardware. Though the system works as analog circuits, the input and output signals including training results are all available in digital format. Therefore, the learned parameters are easily stored and reused after training sessions. The Gaussian Kernel is realized by a new type of circuits utilizing Gilbert multipliers. The proof-of-concept chip was designed and sent to fabrication. Its successful operation was confirmed by transistor level SPICE simulation.
-
an on chip trainable Gaussian Kernel analog support vector machine
International Symposium on Circuits and Systems, 2009Co-Authors: Kyunghee Kang, Tadashi ShibataAbstract:We propose an analog circuit architecture of the Gaussian-Kernel support vector machine having on-chip training capability. Thanks to the hardware-friendly algorithm, the learning function is realized by attaching a small additional circuitry to the SVM classifying hardware. Though the system works as analog circuits, the input and output signals including training results are all available in digital format. Therefore, the learned parameters are easily stored and reused after training sessions. The Gaussian Kernel is realized by a new type of circuits utilizing Gilbert multipliers. The proof-of-concept chip was designed and sent to fabrication. Its successful operation was confirmed by transistor level SPICE simulation.
Witold Pedrycz - One of the best experts on this subject based on the ideXlab platform.
-
high accuracy signal subspace separation algorithm based on Gaussian Kernel soft partition
IEEE Transactions on Industrial Electronics, 2019Co-Authors: Witold Pedrycz, Weike NieAbstract:The separation of signal-and-noise subspaces is a crucial step in many array signal processing applications, since the performance of most high-resolution methods mainly depends on the accuracy of separated signal-and-noise subspaces. If the resulting signal subspace is inaccurately estimated, high-resolution subspace-based algorithms would most likely fail. In this paper, we propose a high-accuracy method to complete the separation of signal-and-noise subspaces. In the developed scheme, eigenvalues of the covariance matrix of the data received by the uniform linear array are first employed to construct an ideal sample space using the Schur product of matrices. Then, the Gaussian Kernel is introduced to map the sample space into a new high-dimensional feature space, where the resulting structure becomes linearly separable. In the sequel, we propose two fuzzy set-based methods to divide the feature space into two subspaces, which correspond to the signal-and-noise subspaces. In this way, the signal subspace becomes separated. Experimental results show that, compared with the results produced by five other commonly used algorithms, the proposed method yields much higher accuracy, especially for low signal-to-noise ratio thresholds and small snapshots.
-
Gaussian Kernel based fuzzy rough sets model uncertainty measures and applications
International Journal of Approximate Reasoning, 2010Co-Authors: Lei Zhang, Degang Chen, Witold PedryczAbstract:Kernel methods and rough sets are two general pursuits in the domain of machine learning and intelligent systems. Kernel methods map data into a higher dimensional feature space, where the resulting structure of the classification task is linearly separable; while rough sets granulate the universe with the use of relations and employ the induced knowledge granules to approximate arbitrary concepts existing in the problem at hand. Although it seems there is no connection between these two methodologies, both Kernel methods and rough sets explicitly or implicitly dwell on relation matrices to represent the structure of sample information. Based on this observation, we combine these methodologies by incorporating Gaussian Kernel with fuzzy rough sets and propose a Gaussian Kernel approximation based fuzzy rough set model. Fuzzy T-equivalence relations constitute the fundamentals of most fuzzy rough set models. It is proven that fuzzy relations with Gaussian Kernel are reflexive, symmetric and transitive. Gaussian Kernels are introduced to acquire fuzzy relations between samples described by fuzzy or numeric attributes in order to carry out fuzzy rough data analysis. Moreover, we discuss information entropy to evaluate the Kernel matrix and calculate the uncertainty of the approximation. Several functions are constructed for evaluating the significance of features based on Kernel approximation and fuzzy entropy. Algorithms for feature ranking and reduction based on the proposed functions are designed. Results of experimental analysis are included to quantify the effectiveness of the proposed methods.