Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning最新文献

An Efficient and Robust Aggregation Algorithm for Learning Federated CNN 一种高效鲁棒的联邦CNN学习聚合算法

Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning Pub Date : 2020-10-22 DOI: 10.1145/3432291.3432303

Yanyang Lu, Lei Fan

{"title":"An Efficient and Robust Aggregation Algorithm for Learning Federated CNN","authors":"Yanyang Lu, Lei Fan","doi":"10.1145/3432291.3432303","DOIUrl":"https://doi.org/10.1145/3432291.3432303","url":null,"abstract":"Federated learning is a privacy-protected way of decentralized machine learning. In a Federated Learning system, the server uses the aggregation algorithm to obtain a global model from clients. The traditional aggregation method called FedAvg is a simple arithmetic average, but it has never been proved efficient. Moreover, there are many potential attacks and network instability in Federated Learning systems. This paper aims to achieve two goals by replacing the server-side aggregation strategy: 1) Accelerate global model's convergence speed 2) Make global model reliable when facing network instability and offline attacks. Considering different clients' contributions have different impacts, we try to use gaussian distribution to weight clients' potential contributions. To make the aggregation process more modality to different neural network architecture, we try to solve the above problems on Convolution Neural Network as a representation. To work well on different functional units in neural networks, we also propose layer-wise optimizing steps. We have experimented on some representative tasks of Federated Learning, and the results show that our method exceeds FedAvg a lot on convergence speed. When simulating attack situations, our algorithm could be proved to maintain a reliable global model.","PeriodicalId":126684,"journal":{"name":"Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127879418","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 11

Comparison of Holt-Winters and ARIMA Models for Hydropower Forecasting in Guangxi 霍尔特-温特斯模型与ARIMA模型在广西水电预报中的比较

Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning Pub Date : 2020-10-22 DOI: 10.1145/3432291.3432307

Yuan Lei, Pingwen Xue, Yanxing Li

{"title":"Comparison of Holt-Winters and ARIMA Models for Hydropower Forecasting in Guangxi","authors":"Yuan Lei, Pingwen Xue, Yanxing Li","doi":"10.1145/3432291.3432307","DOIUrl":"https://doi.org/10.1145/3432291.3432307","url":null,"abstract":"Nowadays, electrical power supply has had in the main role in developing countries, so the hydropower attaches great importance to all countries, in order to solve the electricity problems of industrial and agricultural production and people's daily life. Hydropower is a clean energy source that is more seasonal than other forms of power generation. Forecasting the hydropower generation is a challenging problem, and time series forecasting has been considered as an effective forecasting method. However, under the principle of minimum forecast variance, the longer the forecast time, the greater the variance of the forecast value, so time series data are only suitable for short-term forecasting. This paper used the Holt-Winters additive model and the ARIMA model for solving the shortage of time series forecasting to forecast Guangxi's hydropower generation for the next five months from January 2010 to February 2020, and the forecast results is compared with the real data from the National Bureau of Statistics of China. The result show that ARIMA model is much less accurate than Holt-Winters' additive model in when analysis data with seasonal fluctuations. This paper, can provide some information for energy managers and policymakers.","PeriodicalId":126684,"journal":{"name":"Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122985292","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 3

Information-aware Message Passing Neural Networks for Graph Node Classification 面向图节点分类的信息感知消息传递神经网络

Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning Pub Date : 2020-10-22 DOI: 10.1145/3432291.3432294

Li Zhou, Chunbo Jia, Jianfeng Zhang

{"title":"Information-aware Message Passing Neural Networks for Graph Node Classification","authors":"Li Zhou, Chunbo Jia, Jianfeng Zhang","doi":"10.1145/3432291.3432294","DOIUrl":"https://doi.org/10.1145/3432291.3432294","url":null,"abstract":"In this paper, we present an information-aware message passing neural network (MPNN) to improve the effectiveness of weight calculation in the message passing process for node classification problems on a graph. Related researches are theoretically analyzed to introduce a special information interchanging scheme with precise weight generation. Compared to fix weighted message passing in convolutional networks (GCN) and neural network computed message coefficient in graph attention networks (GAT), the proposed information-aware message passing neural network takes advantage of node feature and its hidden representations, which indicate the amount of information should be passed and how to absorb the message between node neighbours. This exact message-passing method with self-learning for weight on each dimension of feature achieves better accuracy for node classification tasks in graph neural networks. The proposed method is evaluated by semi-supervised graph node classification task on citation network dataset. Obtained results show the correctness of the proposed method which are more accurate than that of GCN and GAT result.","PeriodicalId":126684,"journal":{"name":"Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130568995","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

Deep Learning for Three Types of Keratitis Classification based on Confocal Microscopy Images 基于共聚焦显微镜图像的三种角膜炎分类的深度学习

Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning Pub Date : 2020-10-22 DOI: 10.1145/3432291.3432310

Xinming Zhang, Gang Ding, C. Gao, Chao-Lei Li, Bing-liang Hu, Chenming Zhang, Quan Wang

引用次数: 3

An Instance Segmentation approach to Food Calorie Estimation using Mask R-CNN 基于掩模R-CNN的食物卡路里估计实例分割方法

Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning Pub Date : 2020-10-22 DOI: 10.1145/3432291.3432295

Parth Poply, A. J.

引用次数: 8

An Interference Cancellation Method Using Fixed Beamformer and Adaptive Filter in Car Environment 一种基于固定波束形成器和自适应滤波器的汽车环境干扰消除方法

Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning Pub Date : 2020-10-22 DOI: 10.1145/3432291.3432299

Xi Liu, Xiaoxia Yang

{"title":"An Interference Cancellation Method Using Fixed Beamformer and Adaptive Filter in Car Environment","authors":"Xi Liu, Xiaoxia Yang","doi":"10.1145/3432291.3432299","DOIUrl":"https://doi.org/10.1145/3432291.3432299","url":null,"abstract":"In-car speech interactive system usually requires to selectively be controlled by one or more person in the vehicle. Beamforming is often introduced to deal with this case by using microphone array. However, most of the beamforming methods like fixed beamformer, Minimum Variance Distortionless Response (MVDR) and Generalized Sidelobe Canceller (GSC) show limited performance when two or more passengers speak simultaneously. Speech recognition pass rate will be severely degraded by the residual interference. And in car environment, speaker's location is previously known and relatively fixed, so an optimized interference cancellation method is proposed to separate desired signal from interferences. The proposed method consists of two parts, beamforming by using multiple fixed beamformer and optimized joint adaptive filter. The proposed beamforming method can provide better reference signal with less desired signal component compared with conventional beamforming method. Thus, by utilizing a powerful adaptive filter, residual noise and interference can be further suppressed with acceptable speech distortion. Simulation and real car experimental results show significant improvement on signal to noise plus interference ratio (SINR) and character error rate (CER).","PeriodicalId":126684,"journal":{"name":"Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning","volume":"257 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133365886","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning 2020年第三届信号处理与机器学习国际会议论文集

Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning Pub Date : 2020-10-22 DOI: 10.1145/3432291

引用次数: 0

An Automatic Scheme for Optimizing the Size of Deep Networks 一种自动优化深度网络大小的方案

Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning Pub Date : 2020-10-22 DOI: 10.1145/3432291.3432293

Wenting Ma, Zhipeng Zhang, Qingqing Xu, Wai Chen

{"title":"An Automatic Scheme for Optimizing the Size of Deep Networks","authors":"Wenting Ma, Zhipeng Zhang, Qingqing Xu, Wai Chen","doi":"10.1145/3432291.3432293","DOIUrl":"https://doi.org/10.1145/3432291.3432293","url":null,"abstract":"Large-scale datasets and complex architectures promote the development of CNN models. Although with stronger representation power, larger CNNs are more resource-hungry, which makes it difficult to deploy on resource-constrained Internet of Things (IoT) devices. Another serious challenge occurring with larger CNNs is their susceptibility to overfit with a small training dataset. In this paper, we ask the question: can we find an optimized compact model for a particular data set? We propose a novel scheme to optimize the model size so as to obtain a compact model instead of a larger model. The optimized model achieves higher accuracy than the widely used deeper model. In addition, it decreases the run-time memory and reduces the number of computing operations. This is achieved by applying Minimum Description Length (MDL) to find the optimal size of the model for a particular data set mathematically. MDL--the information-theoretic model selection principle assumes that the simplest, most compact representation model is the best model and most probable explanation of the data. We call our approach OptSize, model size is automatically identified, yielding compact models with comparable accuracy. We empirically demonstrate the effectiveness of our approach with several state-of-the-art CNN models, including VGGNet, ResNet on various image classification datasets. The result shows that compact nets obtained by our proposed method perform better to complex, well-engineered, deeper convolutional architectures.","PeriodicalId":126684,"journal":{"name":"Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128364642","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multi-target Positioning Method of Passive MIMO Radar based on DBSCAN 基于DBSCAN的无源MIMO雷达多目标定位方法

Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning Pub Date : 2020-10-22 DOI: 10.1145/3432291.3432292

Lijun Wang, Yu Han, Li Wang, Yiqi Chen, Wanchun Li

引用次数: 0

The Evolution and Emerging Trends of Cloud Computing Adoption Research: Visual Analysis of CiteSpace Based on WOS Papers 云计算应用研究的演变与新趋势——基于WOS论文的CiteSpace可视化分析

Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning Pub Date : 2020-10-22 DOI: 10.1145/3432291.3433641

Ge Zhang, Yun Chen, Gaoyong Li

{"title":"The Evolution and Emerging Trends of Cloud Computing Adoption Research: Visual Analysis of CiteSpace Based on WOS Papers","authors":"Ge Zhang, Yun Chen, Gaoyong Li","doi":"10.1145/3432291.3433641","DOIUrl":"https://doi.org/10.1145/3432291.3433641","url":null,"abstract":"With the continuous promotion and application of cloud computing, the issue of its adoption has increasingly attracted widespread attention from the theoretical and practical circles at home and abroad. Therefore, it is of great significance to summarize and analyze the research status and hotspots of cloud computing adoption and promote the subsequent research on cloud computing adoption. This paper uses CiteSpace, a visualized literature analysis software, to draw core authors, core journals, key words cluster and other knowledge maps in the field of cloud computing adoption based on 668 documents related to cloud computing adoption from 2011 to 2019 included in the Web of Science database. The research status and trend of cloud computing adoption are visualized and analyzed. The results of the study indicate that the attention paid to cloud computing adoption is gradually increasing. It shows that the research on the security, risk and other technical characteristics of cloud computing, the research on the influencing factors of cloud computing adoption and user behavior, and the research on the cost and performance of cloud computing are the three main research contents in the field of cloud computing adoption. The research hotspots in the field of cloud computing adoption have experienced the evolution process from specific research issues such as cloud computing business and technology, to the application and implementation of cloud computing in enterprises, to the integration of cloud computing and other emerging technologies.","PeriodicalId":126684,"journal":{"name":"Proceedings of the 2020 3rd International Conference on Signal Processing and Machine Learning","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115543926","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 4