检索结果-内蒙古大学图书馆

2021 5th International Conference on Computer Science and Artificial Intelligence

作者： Yawei Zhang Fangtao Ren Xi Liu Fan Zhang School of computer and information engineering Henan University China Henan Key Laboratory of Big Data Analysis and Processing Henan University China

ISBN: (纸本)9781450384155

DNA methylation is of great significance to the diagnosis, treatment and disease prediction of hepatocellular carcinoma (HCC). The commonly used DNA methylation microarrays have high data dimensions, and different CpG sites detected may map to the same gene. To extract more effective features for HCC disease prediction, this study uses a linear regression model that integrates TCGA database methylation data and gene expression data, based on the DNA methylation microarray data of the GEO database (GSE113017), the corresponding gene expression data was predicted and the differentially expressed genes (3766) with significant differences were screened out, which was used as the feature of the data set. Constructing an Artificial Neural Network (ANN) to train a machine learning model for HCC disease prediction and perform 10-fold cross-validation. The resulting model has an accuracy of 95.1% for HCC disease prediction based on DNA methylation microarray data. Compared with other HCC prediction methods, this method has better performance. Then analyze the differentially expressed genes with protein-protein interaction network (PPI network), and use the top five connected genes in the network as hub genes, namely: GNGT2, GNB4, FPR2, CDC20, NMUR1, which can be used as biomarkers for the diagnosis, treatment and prognosis of HCC.

关键词： DNA Methylation Machine learning Artificial neural networks Hepatocellular carcinoma

来源：评论

学校读者我要写书评

暂无评论

Negative Sample Mining based Deep Feature Learning for Kinship Verification 21

Negative Sample Mining based Deep Feature Learning for Kinsh...

引用

Proceedings of the 2021 5th International Conference on Video and Image processing

作者： Xiaoke Zhu Changlong Li Xiaopan Chen Minghao Zheng Yunwei Li Jihui Hu Henan Key Laboratory of Big Data Analysis and Processing Henan University China School of Computer and Information Engineering Henan University China

ISBN: (纸本)9781450385893

Kinship verification is an important and challenging task. The aim of it is to find out whether there is a kin relationship between one pair of facial images. However, there are many challenges for kinship verification. One of the most important challenge is the problem of data imbalance. Specifically, if there exists N positive kinship pairs, we can get N(N-1) negative pairs. Obviously, the number of negative kinship pairs is much larger than the number of positive kinship pairs. How to make full use of positive and negative samples for training is a valuable problem. But most of the existing methods just randomly select the same number of negative sample pairs as the positive sample pairs. What is more, different negative pairs contain different discrimination information. Therefore, finding discriminative sample is also an important problem. In this paper, we propose a simple and effective method, named Negative sample Mining based Deep Feature Learning (NMDFL) to solve the above problems. Specifically, different with most existing methods, we sample N positive kinship pairs and T*N negative pairs, where T>1. Then, they are sent to the train net, and their impacts on network updating are adjusted in proportion to the number of positive and negative samples. At the same time, the corresponding weight of each negative sample is dynamically adjusted according to the training results of each time. Experimental results on the KinFaceW-I and KinFaceW-II datasets proves the effectiveness of our method.

关键词： Deep Feature Learning Image-based Kinship Verification Mining Negative Samples Unbalanced data

来源：评论

学校读者我要写书评

暂无评论

Hyperspectral Image Classification Method Based on Spiking Neural Network

arXiv

引用

arXiv 2024年

作者： Liu, Yang Li, Yahui Li, Rui Zhou, Liming Dang, Lanxue Mu, Huiyu Ge, Qiang Henan Key Laboratory of Big Data Analysis and Processing School of Computer and Information Engineering Henan University Kaifeng475004 China

Convolutional neural network (CNN) performs well in Hyperspectral Image (HSI) classification tasks, but its high energy consumption and complex network structure make it difficult to directly apply it to edge computing devices. At present, spiking neural networks (SNN) have developed rapidly in HSI classification tasks due to their low energy consumption and event driven characteristics. However, it usually requires a longer time step to achieve optimal accuracy. In response to the above problems, this paper builds a spiking neural network (SNN-SWMR) based on the leaky integrate-and-fire (LIF) neuron model for HSI classification tasks. The network uses the spiking width mixed residual (SWMR) module as the basic unit to perform feature extraction operations. The spiking width mixed residual module is composed of spiking mixed convolution (SMC), which can effectively extract spatial-spectral features. Secondly, this paper designs a simple and efficient arcsine approximate derivative (AAD), which solves the non-differentiable problem of spike firing by fitting the Dirac function. Through AAD, we can directly train supervised spike neural networks. Finally, this paper conducts comparative experiments with multiple advanced HSI classification algorithms based on spiking neural networks on six public hyperspectral data sets. Experimental results show that the AAD function has strong robustness and a good fitting effect. Meanwhile, compared with other algorithms, SNN-SWMR requires a time step reduction of about 84%, training time, and testing time reduction of about 63% and 70% at the same accuracy. This study solves the key problem of SNN based HSI classification algorithms, which has important practical significance for promoting the practical application of HSI classification algorithms in edge devices such as spaceborne and airborne devices. © 2024, CC BY.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Fast Parallel Extraction Method of Normalized Vegetation Index

Fast Parallel Extraction Method of Normalized Vegetation Ind...

引用

International Conference on Computer Science & Education (ICCSE)

作者： Xianyu Zuo Tongyuan Qi Baojun Qiao Zhitao Deng Qiang Ge Key Laboratory of Big Data Analysis and Processing School of Computer and Information Engineering Henan University Kaifeng China

ISBN: (数字)9781728172675

ISBN: (纸本)9781728172682

At present, accelerated processing of remote sensing big data has become an important research topic in the field of remote sensing. Remote sensing image processing based on large-scale clusters currently is the mainstream. However, how to fully tap the computing power of a single computing node in the cluster has become issues that cannot be ignored in the field of remote sensing image processing. Traditional GPU programming is difficult to develop, the development cycle is long, and the requirements for developers are very high. In order to improve the efficiency of GPU programming and shorten the development cycle of parallel programs, Nvidia, Grary, PGI and CAPS jointly launched a new programming standard-OpenAcc. In this paper, OpenAcc-NDVI, as a fast parallel extraction method is used to optimize NDVI algorithm. Based on different computing scenarios, two granularity acceleration models are proposed. It has been verified by multiple experiments that when the data size reaches 10000 * 10000, OpenAcc-NDVI can achieve an acceleration of about 5.3 times. After error analysis, the error of algorithm experiment result is 0, there is no loss of precision. The NDVI algorithm based on OpenAcc has excellent acceleration performance, efficient development process, and high calculation accuracy.

关键词： Acceleration Graphics processing units Histograms Parallel processing Vegetation mapping Remote sensing Arrays

来源：评论

学校读者我要写书评

暂无评论

Location Rating Based on Graph Neural Network 2020

Location Rating Based on Graph Neural Network

引用

Proceedings of the 2020 International Conference on Cyberspace Innovation of Advanced Technologies

作者： Fengbin Zheng Li Ding Lanxue Dang Henan Key Laboratory of Big Data Analysis and Processing School of Computer and Information Engineering Henan University Kaifeng PR China

ISBN: (纸本)9781450387828

With the rapid development of social networks, people are increasingly interested in sharing their location. Location recommendation has become an important personalization service for location-based social networks (LBSN). Location rating is one of the important tools. However, location-based social networks contain multidimensional network structures and node information. Existing approaches mostly focus on network structures that utilize one of these dimensions, which makes it difficult to efficiently aggregate information from multiple dimensions simultaneously. To overcome these difficulties, one of the recent approaches is the social recommendation based on graph neural networks (GNNs). Based on this, this paper proposes a graph neural network framework for location rating. Graph neural networks are highly inductive and can efficiently aggregate network structure and node information. In particular, the method not only aggregates homogeneous social networks composed of user and heterogeneous bipartite graphs composed of user-location, but also constructs location networks using location sequences to propose an aggregation model for location networks. Experiments are conducted on two datasets, and the results show that the method improves on the root mean square error (RMSE) and mean absolute error (MAE) metrics.

关键词： Graph neural network Location rating Location based social networks

来源：评论

学校读者我要写书评

暂无评论

Lightweight Fine-Tuning a Pretrained Protein Language Model for Protein Secondary Structure Prediction

SSRN

引用

SSRN 2023年

作者： Yang, Wei Liu, Chun Li, Zheng Henan Key Laboratory of Big Data Analysis and Processing Henan Engineering Laboratory of Spatial Information Processing School of Computer and Information Engineering Henan University Kaifeng475004 China

Pretrained large-scale protein language models, such as ESM-1b and ProtTrans, are becoming the fundamental infrastructure for various protein-related biological modeling tasks. Existing works use mainly pretrained protein language models in feature extraction. However, the knowledge contained in the embedding features directly extracted from a pretrained model is task-agnostic. To obtain task-specific feature representations, a reasonable approach is to fine-tune a pretrained model based on labeled datasets from downstream tasks. To this end, we investigate the fine-tuning of a given pretrained protein language model for protein secondary structure prediction tasks. Specifically, we propose a novel end-to-end protein secondary structure prediction framework involving the lightweight fine-tuning of a pretrained model. The framework first introduces a few new parameters for each transformer block in the pretrained model, then updates only the newly introduced parameters, and then keeps the original pretrained parameters fixed during training. Extensive experiments on seven test sets, namely, CASP12, CASP13, CASP14, CB433, CB634, TEST2016, and TEST2018, show that the proposed framework outperforms existing predictors and achieves new state-of-the-art prediction performance. Furthermore, we also experimentally demonstrate that lightweight fine-tuning significantly outperforms full model fine-tuning and feature extraction in enabling models to predict secondary structures. Further analysis indicates that only a few top transformer blocks need to introduce new parameters, while skipping many lower transformer blocks has little impact on the prediction accuracy of secondary structures. Our code and datasets are publicly available at https://***/fengtuan/LIFT_SS. © 2023, The Authors. All rights reserved.

关键词： Proteins

来源：评论

学校读者我要写书评

暂无评论

Spatial fuzzy C-means clustering and deep belief network for change detection in synthetic aperture radar images 5

Spatial fuzzy C-means clustering and deep belief network for...

引用

IET International Radar Conference 2020, IET IRC 2020

作者： Qi, Wenwen Wu, Lin Guo, Zhengwei Huang, Dan College of Computer and Information Engineering Henan University Kaifeng475004 China Henan Key Laboratory of Big Data Analysis and Processing Henan University Kaifeng475004 China Henan Engineering Research Center of Intelligent Technology and Application Henan University Kaifeng475004 China College of Environment and Planning Henan University Kaifeng475004 China Department of Laboratory and Equipment Management Henan University Kaifeng475004 China

ISBN: (纸本)9781839535635

In this study, spatial fuzzy c-means (SFCM) clustering and deep belief network (DBN) method is presented for change detection in SAR images. There are three primary steps of this approach, they are given as follows: 1) Difference Image (DI) is firstly obtained by calculating the similarity of gray levels between multitemporal SAR images. 2) SFCM clustering is used to analyze DI, which combines the spatial information factor and the fuzzy cluster membership function, thus can effectively eliminate the interference of speckle noise. 3) DBN, a powerful data processing methodology that integrates deep learning techniques to extract underlying features and structures of complicated data, is trained to classify pixels to be changed or unchanged. The effectiveness and robustness of the proposed method are verified by experimental results with the Ottawa dataset. © 2020 Institution of Engineering and Technology. All rights reserved.

关键词： Membership functions

来源：评论

学校读者我要写书评

暂无评论

Sonar Target Detection Based on a Dual Channel Attention Convolutional Network 12

Sonar Target Detection Based on a Dual Channel Attention Con...

引用

12th International Conference on Information Science and Technology, ICIST 2022

作者： Liu, Yang Wang, Ruiyi Cao, Kejing Wang, Jiuhao Shi, Zezhao Wang, Yadi Zhou, Yi Shenzhen Research Institute Henan University Henan Province Engineering Research Center of Spatial Information Processing College of Computer and Information Engineering Kaifeng China College of Computer and Information Engineering Henan University Henan Key Laboratory of Big Data Analysis and Processing Kaifeng China College of Computer and Information Engineering Henan University Henan Province Engineering Research Center of Spatial Information Processing Kaifeng China College of Artificial Intelligence Henan University Shenzhen Research Institute Henan University Zhengzhou China

ISBN: (数字)9781665485821

ISBN: (纸本)9781665485821

Due to the complexity and diversity of underwater environment, high-precision and fast target detection is a scientific problem in underwater acoustic information extraction, especially the underwater target detection of sonar images still has a technical bottleneck. With the development of intelligent detection technology, as the state of the art model, target detection model based on deep neural network adopts different scale feature extraction mechanism, which is easy to generate false alarm for important targets and difficult to overcome the contradiction between false detection and missed detection. The attention mechanism can fully learn the features of the target and improve the accuracy of target detection. Considering the characteristics of seabed exploration task and underwater target, we propose a deep convolution network based on dual channel attention mechanism (DCNet), This model can strengthen the features of the target of interest while weakening the irrelevant background noise information, so as to improve the detection accuracy of the target and enhance the detection ability of the underwater target. The experimental results show that the average accuracy of the dual channel attention mechanism can achieve higher accuracy than the original model, and is superior to other target detection models in accuracy and performance. This research has important practical significance for improving the task of underwater target detection of sonar images and has a wide range of engineering application prospects in the detection of underwater acoustic systems. © 2022 IEEE.

关键词： Underwater acoustics

来源：评论

学校读者我要写书评

暂无评论

Influence of Elevation and Orbit Interpolation on the Accuracy of R-D Location Model

Influence of Elevation and Orbit Interpolation on the Accura...

引用

Asian and Pacific Conference on Synthetic Aperture Radar (APSAR)

作者： Minghu Fan Xianyu Zuo Junfeng Tian Key Laboratory of Big Data Analysis and Processing of Henan Province College of Computer Science and Information Engineering Henan University Kaifeng China

ISBN: (数字)9781728129129

ISBN: (纸本)9781728129136

R-D model is the most commonly location model in geometric processing of SAR images. How to improve its location accuracy has been one of the focuses in application. Based on the in-depth analysis of the solving process of R-D equation, two kinds of influencing factors are selected to study. They are elevation and satellite orbit interpolation method. Then, aiming at the pixel location of high-resolution spaceborne SAR images, three kinds of DEM and two orbit interpolation algorithms were studied. The experimental results show that these two kinds of factors mainly affect the location accuracy of range direction, both the horizontal and vertical resolution of elevation have great influence on location accuracy, the quartic polynomial orbit interpolation can meet requirements of high accuracy location, and further improvement of the location accuracy need other auxiliary means.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Parallel Geometric Correction for Single Spaceborne SAR Image

Parallel Geometric Correction for Single Spaceborne SAR Imag...

引用

Asian and Pacific Conference on Synthetic Aperture Radar (APSAR)

ISBN: (数字)9781728129129

ISBN: (纸本)9781728129136

The fast processing of SAR images is of great significance for SAR applications, and parallel computing is the most common solution. The paper studied the parallel computing algorithm for the geometric correction of single spaceborne SAR image, which is based on a classical serial geometric correction algorithm. After fully analyzing the parallel features of the algorithm, a CPU-oriented and a GPU-oriented global parallel correction algorithm were designed. Then, two spaceborne SAR images were used to test the speedup performance of the two algorithms. The experimental results show that the two parallel algorithms can greatly improve the speed of the geometric correction of spaceborne SAR image, and the performance of CPU-oriented parallel algorithm is mainly related to the number of CPU/cores, while the performance of GPU-oriented parallel algorithm is mainly related to the data granularity, and the latter has very high speedup ratio and is an efficient method for geometric correction of spaceborne SAR image.

关键词：

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：