5,280 Hits in 9.7 sec

A Spatial-Temporal Decomposition Based Deep Neural Network for Time Series Forecasting [article]

Reza Asadi, Amelia Regan
2019 arXiv   pre-print
In preprocessing, a time series decomposition method is applied to separately feed short-term, long-term and spatial patterns into different components of a neural network.  ...  In neural network architecture, each kernel of a multi-kernel convolution layer is applied to a cluster of time series to extract short-term features in neighboring areas.  ...  Fully connected Long-Short Term Memory neural network (LSTM) is capable of capturing long-term temporal patterns.  ... 
arXiv:1902.00636v1 fatcat:5y45pzhzcvfptprelhqncbnckm

A Review of Deep Learning Research

2019 KSII Transactions on Internet and Information Systems  
This paper introduces deep learning techniques from various aspects, including common models of deep learning and their optimization methods, commonly used open source frameworks, existing problems and  ...  Firstly, we introduce the applications of deep learning; Secondly, we introduce several common models of deep learning and optimization methods; Thirdly, we describe several common frameworks and platforms  ...  The forget gate is to make the recurrent neural network "forget" information that was not used before. LSTM can more naturally remember the input long before a long time.  ... 
doi:10.3837/tiis.2019.04.001 fatcat:tefkvk3fvvanbkzwmjn44eoxsu

Optimizing Convolution Neural Network on the TI C6678 multicore DSP

Guozhao Zeng, Xiao Hu, Yueyue Chen, X. Lei, S. Cai, Y. Yu, H. Li
2018 MATEC Web of Conferences  
Convolutional Neural Networks (CNNs) have become the most advanced algorithms for deep learning. They are widely used in image processing, object detection and automatic translation.  ...  The efficiency of the improved convolution operation has increased by tens of times.  ...  The CNN can be regarded as a special multi-layer perceptron or feedforward neural network.  ... 
doi:10.1051/matecconf/201824603044 fatcat:wqijomqzfzds7ashlpzlslipsa

Research on Application of Sports Training Performance Prediction Based on Convolutional Neural Network

Yunlong Li, HaoZhen Zhao, JiaYu Gao, Osamah Ibrahim Khalaf
2022 Computational and Mathematical Methods in Medicine  
Moreover, this paper calculates the regional similarity of sports training images in the fully connected layer of the convolutional neural network and introduces the local linear weighting method for analysis  ...  Finally, this paper combines the convolutional neural network algorithm to construct a sports training performance prediction system to improve the effect of sports training and design experiments to verify  ...  feasibility in time series.  ... 
doi:10.1155/2022/7295833 pmid:35371283 pmcid:PMC8975655 fatcat:5y533qsmerhgvfglwa2rtcjmq4

Parallel Sequence-Channel Projection Convolutional Neural Network for EEG-Based Emotion Recognition

Lili Shen, Wei Zhao, Yanan Shi, Tianyi Qin, Bingzheng Liu
2020 IEEE Access  
Therefore, complete contextual relevance can be obtained via the length-synchronized convolutional kernel.  ...  Song et al. applied a dynamical graph convolutional neural network (DGCNN) to analyze EEG data [17] . Li et al. proposed a hierarchical convolutional neural network [18] .  ... 
doi:10.1109/access.2020.3039542 fatcat:s5lfw6uybvcw5cyymub22axa5q

Minutely Active Power Forecasting Models Using Neural Networks

Dimitrios Kontogiannis, Dimitrios Bargiotas, Aspassia Daskalopulu
2020 Sustainability  
After training and testing long short-term memory (LSTM) network variants, a convolutional neural network (CNN), and a multi-layer perceptron (MLP), we observed that the latter performed better on the  ...  In this study we compared the baseline performance and structure of different types of neural networks on residential energy data by formulating a suitable supervised learning problem, based on real world  ...  Prediction evaluation on 150 data points for (a) Long Short-Term Memory (LSTM) network, (b) Stacked LSTM, (c) Bidirectional LSTM, (d) 1-Dimensional Convolutional Neural Network (1D CNN), (e) Multilayer  ... 
doi:10.3390/su12083177 fatcat:a7rlaf436fdk3aq7tfhzt54x3a

Seizure Prediction Using Multi-View Features and Improved Convolutional Gated Recurrent Network

Lihan Tang, Ning Xie, Menglian Zhao, Xiaobo Wu
2020 IEEE Access  
The PLVs range from 0 to 1, where "0" means two signals are completely asynchronized, and "1" means two signals are fully synchronized.  ...  For a given time series {x k }, the scale-dependent measure µ s,m0 is defined as a root-mean-square fluctuation of the integrated response time series y m around a polynomial trendŷ m,λ of order λ within  ... 
doi:10.1109/access.2020.3024580 fatcat:wwlyaciviverjexkjvwfylty5a

Adaptive Multi-receptive Field Spatial-Temporal Graph Convolutional Network for Traffic Forecasting [article]

Xing Wang
2021 arXiv   pre-print
, and (3) introducing an extra decoder based on a fully connected deep network to conquer the error propagation challenge with multi-step forecasting.  ...  We propose a novel deep learning network architecture, Adaptive Multi-receptive Field Spatial-Temporal Graph Convolutional Networks (AMF-STGCN), to model the traffic dynamics of mobile base stations.  ...  AMF-STConv block output Layer Then, we use a two-layer fully connected neural network to generate the output of AMF-STConv block. We first reshape Ao into Ao ∈ R N ×T ×K× (Co×B) .  ... 
arXiv:2111.00724v1 fatcat:tr6i5ut6bbgf5awjkcc75jce3m

Deep learning for spatio‐temporal supply and demand forecasting in natural gas transmission networks

Milena Petkovic, Thorsten Koch, Janina Zittel
2021 Energy Science & Engineering  
K E Y W O R D S convolutional neural networks, deep learning, natural gas forecasting, spatio-temporal model 2 |  ...  We propose a deep learning model based on spatio-temporal convolutional neural networks (DLST) to tackle the problem of gas flow forecasting in a complex high-pressure transmission network.  ...  Graph based CNN are often used together with recurrent neural networks (RNN) to capture both spatial and temporal dependencies in graph structured time series data.  ... 
doi:10.1002/ese3.932 fatcat:xrxkgump5nhm3ds6iy7mme6z74

Multimodal Sensor Motion Intention Recognition Based on Three-Dimensional Convolutional Neural Network Algorithm

Mofei Wen, Yuwei Wang, Syed Hassan Ahmed
2021 Computational Intelligence and Neuroscience  
We divide the target data into multiple segments and use a three-dimensional convolutional neural network to extract the short-term spatiotemporal features.  ...  The three types of features of the same segment are fused together and input into the LSTM network for time-series modeling to further fuse the features to obtain multimodal long-term spatiotemporal features  ...  Convolutional neural network (CNN) is a special deep feedforward neural net- work.  ... 
doi:10.1155/2021/5690868 pmid:34188674 pmcid:PMC8192210 fatcat:enj3jdfaavh7vcioawkykvnafm

Research on Damage Evaluation of Radar Target Based on Deep Learning

Pengcheng Xu, Jiansheng Shu
2019 IOP Conference Series: Materials Science and Engineering  
; train the damage assessment network to make the network have strong experts Experience has led to a neural network with expert experience.  ...  Determine the network parameters and network node structure characteristics according to the characteristics of damage information; use the appropriate software and tools to build a modular network model  ...  This paper focuses on convolutional neural networks and restricted Boltzmann machines [1] . (1) Convolutional neural network CNN [2] [3] is an extension of neural network.  ... 
doi:10.1088/1757-899x/569/5/052050 fatcat:6qlwjjdvdfai7ivyzfqrjncbri

A New Hybrid Teaching Model for a Psychology Course

Shuyang Hua, Fulian Liu
2021 International Journal of Emerging Technologies in Learning (iJET)  
In addition, a new performance prediction model based on graph convolution neural network is proposed decision tree data mining methods are used to predict student performance, by extracting user information  ...  and high-dimensional information of the knowledge graph, this mode is conducive to platform managers to understand students' conditions in time and improve teaching quality.  ...  Based on graph convolution neural network, this article proposes a feasible method to predict students' learning situation.  ... 
doi:10.3991/ijet.v16i03.20457 fatcat:7c2fnka335ccndrq53k3pm4zae

A Novel Multichannel Dilated Convolution Neural Network for Human Activity Recognition

Yingjie Lin, Jianning Wu
2020 Mathematical Problems in Engineering  
A novel multichannel dilated convolution neural network for improving the accuracy of human activity recognition is proposed.  ...  and not to consider the use of the pooling layers that are used in the traditional convolution with dilated convolution.  ...  Based on current research deficiencies, this paper proposes a novel multichannel dilated convolution neural network (MDCNN). e model can get a larger receptive field to extract global features of long-time  ... 
doi:10.1155/2020/5426532 fatcat:ekfnqzgvszgytj34ftarq3c2om

Early-Stage Gas Identification Using Convolutional Long Short-Term Neural Network with Sensor Array Time Series Data

Kai Zhou, Yixin Liu
2021 Sensors  
In this context, a tailored approach so-called convolutional long short-term memory (CLSTM) neural network is developed to perform the identification task incorporating temporal characteristics within  ...  subsequent gas identification using time-series data during the transitional phase before reaching equilibrium.  ...  In this context, we develop a convolutional long short-term memory (CLSTM) neural network with the convolutional layers embedded to rapidly and automatically extract the low-level features from the time-series  ... 
doi:10.3390/s21144826 fatcat:e7mlx5667zfchaz4m6n2c5mipy

STCGAT: Spatial-temporal causal networks for complex urban road traffic flow prediction [article]

Wei Zhao, Shiqi Zhang, Bing Zhou, Bei Wang
2022 arXiv   pre-print
Existing approaches usually use fixed traffic road network topology maps and independent time series modules to capture Spatial-temporal correlations, ignoring the dynamic changes of traffic road networks  ...  The model dynamically captures the spatial dependence of the traffic network through a Graph Attention Network(GAT) and then analyzes the causal relationship of the traffic data using our proposed Causal  ...  convolutional network is mainly composed of a combination of Bi-directional Long-short Term Memory(BiLSTM) [11] and TCN; 3) Prediction layer: Prediction results are output using a fully connected neural  ... 
arXiv:2203.10749v1 fatcat:k5aw6iqhofdu7bep35fymp3f4m
« Previous Showing results 1 — 15 out of 5,280 results