检索结果-内蒙古大学图书馆

Fuzzy deep learning based urban traffic incident detection

COGNITIVE SYSTEMS RESEARCH 2018年 50卷 206-213页

作者： El Hatri, Chaimae Boumhidi, Jaouad Sidi Mohamed Ben AbdEllah Univ Fac Sci Dhar Mahraz Comp Sci Dept LIIAN Lab Fes 30000 Morocco

Traffic incident detection (TID) is an important part of any modern traffic control because it offers an opportunity to maximise road system performance. For the complexity and the nonlinear characteristics of traffic incidents, this paper proposes a novel fuzzy deep learning based TID method which considers the spatial and temporal correlations of traffic flow inherently. Parameters of the deep network are initialized using a Stacked Auto-Encoder (SAE) model following a layer by layer pre-training procedure. To conduct the fine tuning step, the back-propagation algorithm is used to precisely adjust the parameters in the deep network. Fuzzy logic is employed to control the learning parameters where the objective is to reduce the possibility of overshooting during the learning process, increase the convergence speed and minimize the error. To find the best architecture of the deep network, we used a separate validation set to evaluate different architectures generated randomly based on the Mean Squared Error (MSE). Simulation results show that the proposed incident detection method has many advantages such as higher detection rate and lower false alarm rate. (C) 2017 Published by Elsevier B.V.

关键词： Automatic traffic incident detection Fuzzy deep learning Deep artificial neural network Stacked auto-encoder model back-propagation algorithm Fuzzy logic controller

来源：评论

学校读者我要写书评

暂无评论

Multi-threaded learning control mechanism for neural networks

引用

FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE 2018年 87卷 16-34页

作者： Polap, Dawid Wozniak, Marcin Wei, Wei Damasevicius, Robertas Silesian Tech Univ Inst Math Kaszubska 23 PL-44100 Gliwice Poland Xian Univ Technol Sch Comp Sci & Engn Xian 710048 Shaanxi Peoples R China Kaunas Univ Technol Software Engn Dept Studentu 50 Kaunas Lithuania

Neural networks are applicable in many solutions for classification, prediction, control, etc. The variety of purposes is growing but with each new application the expectations are higher. We want neural networks to be more precise independently of the input data. Efficiency of the processing in a large manner depends on the training algorithm. Basically this procedure is based on the random selection of weights in which neurons connections are burdened. During training process we implement a method which involves modification of the weights to minimize the response error of the entire structure. Training continues until the minimum error value is reached however in general the smaller it is, the time of weight modification is longer. Another problem is that training with the same set of data can cause different training times depending on the initial weight selection. To overcome arising problems we need a method that will boost the procedure and support final precision. In this article, we propose the use of multi-threading mechanism to minimize training time by rejecting unnecessary weights selection. In the mechanism we use a multi-core solution to select the best weights between all parallel trained networks. Proposed solution was tested for three types of neural networks (classic, sparking and convolutional) using sample classification problems. The results have shown positive aspects of the proposed idea: shorter training time and better efficiency in various tasks. (C) 2018 Elsevier B.V. All rights reserved.

关键词： Multi-threading Neural networks back-propagation algorithm

来源：评论

学校读者我要写书评

暂无评论

Quantum-Inspired Stacked Auto-encoder-Based Deep Neural Network algorithm (Q-DNN)

引用

ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING 2018年第12期43卷 6929-6943页

作者： Patel, Om Prakash Tiwari, Aruna Bagade, Vinay Indian Inst Technol Indore Dept Comp Sci & Engn Simrol Indore India

In this paper, a quantum-inspired stacked auto-encoder-based deep neural network (Q-DNN) learning algorithm is proposed. The proposed Q-DNN uses stacked auto-encoder to form a deep neural network. This quantum computing concept has been used to optimize the learning parameters of the algorithm. To form a deep neural network, the selection of learning algorithms and its parameters is an important criterion for achieving better performance. However, learning capability depends upon the proper selection of learning parameters;improper selection may cause local minima or local maxima problem. To avoid improper selection of learning algorithm parameters, the quantum computing concept has been used, which is characterized by the representation of population dynamic, and evaluation function. It uses qubit for probabilistic representation which gives a large subspace for exploration to find an optimal value of the learning parameters. This concept has been utilized in the formation of stacked auto-encoder-based deep neural network by optimizing learning parameters of the learning algorithm. The proposed Q-DNN is trained and tested on various benchmark datasets such as the BUPA Liver Disorder, Ionosphere, PIMA Indians Diabetes, and the MNIST dataset on three different neural network architectures. Firstly, the deep neural network architecture is formed with one hidden layer, then deep neural network formed with two hidden layers, and finally deep neural network formed with three hidden layers. The proposed Q-DNN achieves promising results in terms of classification accuracy, sensitivity, and specificity in comparison with other approaches.

关键词： Deep neural network Stacked auto-encoder Quantum gates back-propagation algorithm Classification

来源：评论

学校读者我要写书评

暂无评论

Outdoor Path Loss Predictions Based on Extreme Learning Machine

引用

WIRELESS PERSONAL COMMUNICATIONS 2018年第1期99卷 441-460页

作者： Popoola, Segun I. Misra, Sanjay Atayero, Aderemi A. Covenant Univ Dept Elect & Informat Engn Ota Nigeria Atilim Univ Dept Comp Engn Ankara Turkey

In a typical outdoor environment, the propagation of radio waves is usually random in nature, to the extent that the characterization of the wireless channel often becomes very difficult. Several models have been developed to predict the average Received Signal Strength (RSS) for specified distance ranges. However, the use of deterministic models requires high computational efficiency while the prediction results of empirical models may not be as accurate as required. On machine learning approach, the performances of multi-layered feed-forward network models are limited by slow convergence and local minimum, such that a global optimal solution is not guaranteed. In this paper, Extreme Learning Machine (ELM) algorithm is considered in the development of an optimal path loss prediction model for outdoor propagation scenario. Single Hidden Layer Feed-forward Neural Networks (SHLFNNs) are trained and tested with the path loss data that were computed based on the RSS data of a commercial 1800 MHz base station located along Lagos-Badagry highway in Nigeria. The training speed, learning effectiveness, and the generalization ability of Artificial Neural Network back-propagation (ANN-BP) and ELM algorithms are analysed. Experimental results show that ELM models are 140 times faster to train than the ANN-BP models. On prediction accuracy, the outputs of ELM, ANN-BP, Okumura-Hata, and COST-231 models have Root Mean Squared Error (RMSE) values of 2.896, 2.449, 7.456, and 6.116 dB respectively;and regression coefficient (R) values of 0.959, 0.973, 0.935, and 0.935 respectively, when compared to the target variable of the training dataset. When the models were tested with new input data that were excluded from the training process, RMSE values of 4.250, 6.622, 8.732, and 7.087 respectively;and R values of 0.893, 0.876, 0.904, and 0.904 respectively are obtained. In conclusion, the findings of this study confirm that ELM algorithm guarantees an optimal path loss model with fast tr

关键词： Extreme Learning Machine Path loss prediction back-propagation algorithm Feed-forward neural network Radio network planning

来源：评论

学校读者我要写书评

暂无评论

Passive Fault Tolerant Control System Using Feed-forward Neural Network for Two-Tank Interacting Conical Level Control System Against Partial Actuator Failures and Disturbances ⁎

引用

IFAC-PapersOnLine 2019年第14期52卷 141-146页

作者： Himanshukumar R. Patel Vipul A. Shah Dharmsinh Desai University Nadiad-387001 Gujarat India

A novel approach for passive fault-tolerant control (PFTC) system design against actuator faults is proposed. The scheme is based on Feed-forward Neural Network (FFNN) plus conventional PI controller, during the fault occurred into system FFNN will give additional control output to the system according to fault magnitude. The FFNN will trained using back-propagation algorithm. To eliminate the steady-state tracking error, the PI controller is also incorporated. The following fault type and input signals are considered: abrupt, step, sine, and trapezoidal trajectory inputs. The effectiveness and the superiority of the proposed approach are demonstrated using Two-Tank Interacting Conical Level Control System (TTICLCS) example. The simulation performed in MATLAB Simulink platform, also different integral errors like IAE, ISE and IATE are presented to validate the proposed approach.

关键词： Actuator fault back-propagation algorithm conical tank disturbances Feed-forward Neural Network fault-tolerant control PI control

来源：评论

学校读者我要写书评

暂无评论

DeltaFrame-BP: An algorithm Using Frame Difference for Deep Convolutional Neural Networks Training and Inference on Video Data

引用

IEEE TRANSACTIONS ON MULTI-SCALE COMPUTING SYSTEMS 2018年第4期4卷 624-634页

作者： Han, Bing Roy, Kaushik Purdue Univ Dept Elect & Comp Engn W Lafayette IN 47906 USA

Inspired by the success of deep convolutional neural networks (CNNs) with back-propagation (BP) training on large-scale image recognition tasks, recent research efforts concentrated on expending deep CNNs toward more challenging automatized video analysis, such as video classification, object tracking, action recognition and optical flow detection. Video comprises a sequence of images (frames) captured over time in which image data is a function of space and time. Extracting three-dimensional spatial-temporal features from multiple frames becomes a key ingredient for capturing and incorporating appearance and dynamic representations using deep CNNs. Hence, training deep CNNs on video involves significant computational resources and energy consumption due to extended number of frames across the time line of video length. We propose DeltaFrame-BP, a deep learning algorithm, which significantly reduces computational cost and energy consumption without accuracy degradation by streaming frame differences for deep CNNs training and inference. The inherent similarity between video frames due to high fps (frames per second) in video recording helps achieving high-sparsity and low-dynamic range data streaming using frame differences in comparison with raw video frames. According to our simulation, nearly 25 percent energy reduction was achieved in training using the proposed accuracy-lossless DeltaFrame-BP algorithm in comparison with the standard back-propagation algorithm.

关键词： Computer vision deep neural network back-propagation algorithm convolutional neural network energy efficiency

来源：评论

学校读者我要写书评

暂无评论

3-Dimensional Path Planning for Autonomous Underwater Vehicle

3-Dimensional Path Planning for Autonomous Underwater Vehicl...

引用

Conference on OCEANS MTS/IEEE Charleston

作者： Wang, Yiheng Zheng, Yahong Rosa Missouri Univ Sci & Technol Dept ECE Rolla MO 65409 USA Lehigh Univ Dept ECE Bethlehem PA 18015 USA

ISBN: (纸本)9781538648148

This paper proposes a new algorithm for Autonomous Underwater Vehcle (AUV) path planning in 3D space to visit multiple targets using Dubins curves. For a given target-sequence, the 3D path planning is usually solved by two steps: Step 1 projects 3D targets to the X-Y plane and designs a 2D path on this plane;Step 2 maps the 2D path into 3D via interpolation. The proposed new algorithm defines a local coordinate system (LCS) for each pair of targets and designs the 2D Dubins curve in the LCS, then uses the Euler's rotation transformation to convert the 2D dubins curve into the 3D global coordinate system (GCS). Applying the proposed rotation method to a given target sequence and given incoming-outgoing angles yields 3D Dubins path with guaranteed G2 continuity at the joints of two Dubins curves. The proposed method is compared with the interpolation method and Bezier curve method. Computer simulations demonstrate that the proposed algorithm provides better G(2) continuity in 3D space or shorter path lengths than the existing linear or spline interpolation methods and Bezier curves.

关键词： Autonomous Underwater Vehicles AUV Path Planning 3D Dubins Curve Bezier Curve Geometric Continuity back-propagation algorithm

来源：评论

学校读者我要写书评

暂无评论

Through-the-Wall Imaging Using Wi-Fi Signals 12

Through-the-Wall Imaging Using Wi-Fi Signals

引用

12th International Symposium on Antennas, propagation and Electromagnetic Theory (ISAPE)

作者： Zhong, Wei He, W. Kai Wang, Longgang Li, Lianlin Peking Univ Sch Elect Engn & Comp Sci Beijing Peoples R China

ISBN: (纸本)9781538673027

In this work, we investigate, for the first time, the high-resolution three-dimensional (3D) through-the-wall imaging using the everyday wireless communication signals, i.e., Wi-Fi signals, in an indoor environment. We also provide physical insights into such imaging technique, which has been derived from the well-known Huygens principle. Experimentally, the data acquisition has been realized in a synthetic aperture way, and the classical back-propagation algorithm is employed to form the high-resolution imaging. Moreover, we use IEEE 802.11n protocol wireless router working at 2.4GHz, the bandwidth is 20MHz, in our experiments. It can be faithfully expected that such imaging technique can open an exciting new door for various practical applications including safety screening and others.

关键词： through-the-wall Wi-Fi signals back-propagation algorithm

来源：评论

学校读者我要写书评

暂无评论

Beyond back-propagation Learning for Diabetic Detection: Convergence Comparison of Gradient Descent, Momentum and Adaptive Learning Rate 1

Beyond Back-Propagation Learning for Diabetic Detection: Con...

引用

1st International Conference on Informatics and Computational Sciences (ICICoS)

作者： Endah, Sukmawati N. Widodo, Aris P. Fariq, Muhammad L. Nadianada, Shavira I. Maulana, Fadil Univ Diponegoro Dept Informat Comp Sci Semarang Indonesia

ISBN: (纸本)9781538609033

back-propagation (BP) algorithm is one of the classification algorithms in Artificial Neural Network (ANN). back-propagation still has problems with the training process, where learning convergence of gradient descent is very slow. Increase convergence rate can be done with the use of Adaptive Learning Rate and momentum optimization. Previous research that combining both optimization techniques was able to accelerate the training process. Therefore, this research will compare the convergence of gradient descent, gradient descent + momentum, gradient descent + adaptive learning rate and gradient descent + momentum + adaptive learning rate for diabetic detection. Based on the experimental results, it is known that gradient descent has better performance than others in terms of convergence velocity and result of TPR and FPR value. Training algorithm with combining of gradient descent + momentum + adaptive learning rate faster convergent compared to gradient + momentum or gradient + adaptive learning rate.

关键词： back-propagation algorithm gradient descent adaptive learning rate momentum diabetic detection

来源：评论

学校读者我要写书评

暂无评论

Prediction of Solitary Wave Forces on Coastal Bridge Decks Using Artificial Neural Networks

引用

JOURNAL OF BRIDGE ENGINEERING 2018年第5期23卷 4018023.1-4018023.1页

作者： Xu, Guoji Chen, Qin Chen, Jianhua Louisiana State Univ Ctr Computat & Technol Baton Rouge LA 70803 USA Louisiana State Univ Div Comp Sci & Engn Baton Rouge LA 70803 USA Univ Notre Dame NatHazModeling Lab South Bend IN 46556 USA Northeastern Univ Dept Civil & Environm Engn Boston MA 02115 USA Northeastern Univ Dept Marine & Environm Sci Boston MA 02115 USA LA State Univ Div Comp Sci & Engn Baton Rouge LA 70803 USA

This study proposes an alternative and competitive methodology for predicting solitary wave forces on coastal bridge decks using artificial neural networks (ANNs). It is imperative to accurately predict the on-deck wave forces for the design and retrofit of coastal bridges subject to the potential impact of hurricanes and tsunamis. For this purpose, ANNs are used to determine wave loads based on a valid data set. First, the structural, fluid, and wave variables involved in the bridge deck-wave interaction are briefly introduced. The back-propagation network (BPN) wave force prediction model trained using the back-propagation algorithm is highlighted. A data set with 472 evidence cases is prepared based on extensive computational fluid dynamics (CFD) simulations. Three major input variables, the still-water level (SWL), bottom elevation of the girder/superstructure, and wave height, are selected. Then, the procedures of training the ANNs for the vertical and horizontal forces are presented in detail. Finally, the trained network structures with high predictive skills after substantial training are given with a proposed predictive equation for the vertical and horizontal forces. The results showed that the ANN methodology is robust and capable of capturing the underlying physical complexity in the bridge deck-wave interaction.

关键词： Solitary wave forces Coastal bridge decks Artificial neural networks (ANNs) back-propagation algorithm Multilayer perceptron

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：