In order to improve the accuracy of text classification, we present a new convolution neural network model combining keyword and word-meaning transformation. We first preprocess the text and break words, and use sense labeling for semantic keywords and word-meaning transformation. and we divide the texts into two parts---word and word-meaning. Next, we use embedding layer to transform the word and word-meaning into corresponding word embedding. Then, we use improved convoluted neural network to train the model and extract higher-order features of text type data, and use multi-layer perceptron and SoftMax layer to classify the texts to predict the category of each text. Experimental results show that our document classification algorithm can get a high accuracy and the effect of classification of news topic detection gets well.
{"title":"Document Classification Based on semantic and Improved Convolutional Neural Network","authors":"Rong Li, Wei-Bai Zhou, Wei Liu","doi":"10.1145/3417188.3417196","DOIUrl":"https://doi.org/10.1145/3417188.3417196","url":null,"abstract":"In order to improve the accuracy of text classification, we present a new convolution neural network model combining keyword and word-meaning transformation. We first preprocess the text and break words, and use sense labeling for semantic keywords and word-meaning transformation. and we divide the texts into two parts---word and word-meaning. Next, we use embedding layer to transform the word and word-meaning into corresponding word embedding. Then, we use improved convoluted neural network to train the model and extract higher-order features of text type data, and use multi-layer perceptron and SoftMax layer to classify the texts to predict the category of each text. Experimental results show that our document classification algorithm can get a high accuracy and the effect of classification of news topic detection gets well.","PeriodicalId":373913,"journal":{"name":"Proceedings of the 2020 4th International Conference on Deep Learning Technologies","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126946836","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}
A. García, David Barragan-Alcantar, Ivana Collado-Gonzalez, Leonardo Garrido
This paper presents a low-level controller for an unmanned surface vehicle based on Adaptive Dynamic Programming (ADP) and deep reinforcement learning (DRL). The model-based algorithm Back-propagation Through Time and a simulation of the mathematical model of the vessel are implemented to train a deep neural network to drive the surge speed and yaw dynamics. The controller presents successful simulation results validating the feasibility of the proposed strategy and contributes to the diversity of validated applications of ADP and DRL control strategies.
{"title":"Control of an Unmanned Surface Vehicle Based on Adaptive Dynamic Programming and Deep Reinforcement Learning","authors":"A. García, David Barragan-Alcantar, Ivana Collado-Gonzalez, Leonardo Garrido","doi":"10.1145/3417188.3417194","DOIUrl":"https://doi.org/10.1145/3417188.3417194","url":null,"abstract":"This paper presents a low-level controller for an unmanned surface vehicle based on Adaptive Dynamic Programming (ADP) and deep reinforcement learning (DRL). The model-based algorithm Back-propagation Through Time and a simulation of the mathematical model of the vessel are implemented to train a deep neural network to drive the surge speed and yaw dynamics. The controller presents successful simulation results validating the feasibility of the proposed strategy and contributes to the diversity of validated applications of ADP and DRL control strategies.","PeriodicalId":373913,"journal":{"name":"Proceedings of the 2020 4th International Conference on Deep Learning Technologies","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128376381","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}
In this paper, we mainly describe the Internet of things technology involved in teenagers' health monitoring value, methods, steps and matters needing attention. Physical fitness of teenagers continues to decline, our country attaches great importance to the monitoring and management of teenagers' physical health. It should become the trend of physique tests of teenager to use high technology to ensure the effectiveness of information collection, the standardization of processing, the accuracy of the test results data and the scientific nature of physique tests. The rapid development of the Internet of things (hereinafter referred to as "IOT") provides important support for the realization of this trend. With the methods of Literature review and induction-deduction, this paper clarifies 6 core parts of physique tests of Teenagers under the IOT, discusses the intervention paths of this technology, puts forward links requiring improvement so as to enhance the application of IOT in physique tests of teenagers, strengthening physical fitness monitoring, guiding students' exercise, feedback, and goal achievement, and to help improving the physical fitness of teenagers.
{"title":"Monitoring and Management of Teenagers' Physical Health Based on Internet of Things Technology","authors":"Donghai Wu, Ying Ming","doi":"10.1145/3417188.3417190","DOIUrl":"https://doi.org/10.1145/3417188.3417190","url":null,"abstract":"In this paper, we mainly describe the Internet of things technology involved in teenagers' health monitoring value, methods, steps and matters needing attention. Physical fitness of teenagers continues to decline, our country attaches great importance to the monitoring and management of teenagers' physical health. It should become the trend of physique tests of teenager to use high technology to ensure the effectiveness of information collection, the standardization of processing, the accuracy of the test results data and the scientific nature of physique tests. The rapid development of the Internet of things (hereinafter referred to as \"IOT\") provides important support for the realization of this trend. With the methods of Literature review and induction-deduction, this paper clarifies 6 core parts of physique tests of Teenagers under the IOT, discusses the intervention paths of this technology, puts forward links requiring improvement so as to enhance the application of IOT in physique tests of teenagers, strengthening physical fitness monitoring, guiding students' exercise, feedback, and goal achievement, and to help improving the physical fitness of teenagers.","PeriodicalId":373913,"journal":{"name":"Proceedings of the 2020 4th International Conference on Deep Learning Technologies","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129879963","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}
There are some common problems in vocational and technical education for NCO (Non-Commissioned of Officers), such as the low learning enthusiasm of students, weak integration of teaching contents and position requirements, and the simple teaching method, etc. In order to solve these problems, task-driven flipped classroom is attempted to be applied in air conditioning technology and application course, based on in-depth study of course contents, CO students' characteristics and actual job demands. In this method, task is the main line, flipped classroom is used to organize the teaching, the subject status of students is fully reflected, and the learning is improved significantly.
{"title":"Application of Task-driven Flipped Classroom in NCO's Vocational and Technical Education","authors":"Cai-zhen Hong, Lv Jin","doi":"10.1145/3417188.3417213","DOIUrl":"https://doi.org/10.1145/3417188.3417213","url":null,"abstract":"There are some common problems in vocational and technical education for NCO (Non-Commissioned of Officers), such as the low learning enthusiasm of students, weak integration of teaching contents and position requirements, and the simple teaching method, etc. In order to solve these problems, task-driven flipped classroom is attempted to be applied in air conditioning technology and application course, based on in-depth study of course contents, CO students' characteristics and actual job demands. In this method, task is the main line, flipped classroom is used to organize the teaching, the subject status of students is fully reflected, and the learning is improved significantly.","PeriodicalId":373913,"journal":{"name":"Proceedings of the 2020 4th International Conference on Deep Learning Technologies","volume":"155 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127227138","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}
{"title":"Proceedings of the 2020 4th International Conference on Deep Learning Technologies","authors":"","doi":"10.1145/3417188","DOIUrl":"https://doi.org/10.1145/3417188","url":null,"abstract":"","PeriodicalId":373913,"journal":{"name":"Proceedings of the 2020 4th International Conference on Deep Learning Technologies","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133048205","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}
In recent years, benefiting from the introduction of deep learning network algorithms such as Inception, Resnet, and Mobilenet, the accuracy of object classification has been significantly improved, especially for flower classification. Furthermore, with the development of mobile terminals, it becomes common for non-professional people to take photos of wild flowers, which makes flower classification an attractive feature. However, due to the blur effect of photos, it is challenging to achieve high accuracy and robustness in terms of classification. In this paper, we propose a three-step automatic classification scheme based on Inception network. We first preprocess the flower image to filter out blurred images. Then, the images in the training set are segmented by GrabCut, and the flowers are segmented by background to increase the number of samples in the training set. Then, we adopt the Inception-V3 network to extract the features of clear images and perform classification. The results show that the proposed scheme can improve the classification accuracy rate by a maximum of 40.35 %, reaching 97.78 %.
{"title":"Classification of Flowers under Complex Background Using Inception-V3 Network","authors":"Zongliang Gao, Meng Li, Wei Li, Qi Yan","doi":"10.1145/3417188.3417192","DOIUrl":"https://doi.org/10.1145/3417188.3417192","url":null,"abstract":"In recent years, benefiting from the introduction of deep learning network algorithms such as Inception, Resnet, and Mobilenet, the accuracy of object classification has been significantly improved, especially for flower classification. Furthermore, with the development of mobile terminals, it becomes common for non-professional people to take photos of wild flowers, which makes flower classification an attractive feature. However, due to the blur effect of photos, it is challenging to achieve high accuracy and robustness in terms of classification. In this paper, we propose a three-step automatic classification scheme based on Inception network. We first preprocess the flower image to filter out blurred images. Then, the images in the training set are segmented by GrabCut, and the flowers are segmented by background to increase the number of samples in the training set. Then, we adopt the Inception-V3 network to extract the features of clear images and perform classification. The results show that the proposed scheme can improve the classification accuracy rate by a maximum of 40.35 %, reaching 97.78 %.","PeriodicalId":373913,"journal":{"name":"Proceedings of the 2020 4th International Conference on Deep Learning Technologies","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128294147","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}
Credit risk assessment has been thought of as a critical factor in financial companies and banks in the history of development economics. Recently, there has been renewed interest in credit risk assessment using deep learning methods. However, previous studies have not fine-grained dealt with static and dynamic features, which limits their effectiveness. Thus, in this paper, we present a two-stage model using FeedForward Neural Network(FNN) and Recurrent Neural Network(RNN). First, we design the aggregation layer to extract representative information from the static feature at time T. Second, the distinct moment representation constructs the dynamic features of a client. The dynamic features could be learned by the RNN layer. Experimental results on the real-world dataset show its superiority over various baselines.
{"title":"A Two-Stage Dynamic Credit Risk Assessment System","authors":"Rui Li, Shizhe Deng, Jianquan Zhang, Hao He, Yaohui Jin, Jiangang Duan","doi":"10.1145/3417188.3417193","DOIUrl":"https://doi.org/10.1145/3417188.3417193","url":null,"abstract":"Credit risk assessment has been thought of as a critical factor in financial companies and banks in the history of development economics. Recently, there has been renewed interest in credit risk assessment using deep learning methods. However, previous studies have not fine-grained dealt with static and dynamic features, which limits their effectiveness. Thus, in this paper, we present a two-stage model using FeedForward Neural Network(FNN) and Recurrent Neural Network(RNN). First, we design the aggregation layer to extract representative information from the static feature at time T. Second, the distinct moment representation constructs the dynamic features of a client. The dynamic features could be learned by the RNN layer. Experimental results on the real-world dataset show its superiority over various baselines.","PeriodicalId":373913,"journal":{"name":"Proceedings of the 2020 4th International Conference on Deep Learning Technologies","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121351791","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}
In the volume image of brain MRI, the volume of hippocampus is small, the boundary between hippocampus and surrounding tissue is fuzzy, and the two-dimensional semantic segmentation network is difficult to accurately segment. In this paper, an algorithm is proposed which combines deep residual learning and U-net for hippocampus segmentation of brain MRI volume image. It can make full use of the three-dimensional spatial information of MRI image itself, improve the ability of automatic and precise extraction of image features, and achieve high-precision hippocampus segmentation of MRI volume image. Firstly, in order to efficiently utilize 3d contextual information of the image and the solve class imbalance issue, the patches were extracted from brain MRI volume image and put into network. Then, the segmentation model based on the combination of depth residual learning and U-net is used to extract the features of image patches. After that, the upper sampling feature map and the residual learning feature map are fused to get the volume segmentation results. Finally, the detection experiments on ADNI dataset show that DSC (dice similarity coefficient) can reach 0.8915, which is better than the traditional segmentation method.
{"title":"Combining Residual learning and U-Net for Hippocampus Segmentation of Brain MRI Volume Image","authors":"Chao Jia, Changrun Jia, Hailan Yu","doi":"10.1145/3417188.3417191","DOIUrl":"https://doi.org/10.1145/3417188.3417191","url":null,"abstract":"In the volume image of brain MRI, the volume of hippocampus is small, the boundary between hippocampus and surrounding tissue is fuzzy, and the two-dimensional semantic segmentation network is difficult to accurately segment. In this paper, an algorithm is proposed which combines deep residual learning and U-net for hippocampus segmentation of brain MRI volume image. It can make full use of the three-dimensional spatial information of MRI image itself, improve the ability of automatic and precise extraction of image features, and achieve high-precision hippocampus segmentation of MRI volume image. Firstly, in order to efficiently utilize 3d contextual information of the image and the solve class imbalance issue, the patches were extracted from brain MRI volume image and put into network. Then, the segmentation model based on the combination of depth residual learning and U-net is used to extract the features of image patches. After that, the upper sampling feature map and the residual learning feature map are fused to get the volume segmentation results. Finally, the detection experiments on ADNI dataset show that DSC (dice similarity coefficient) can reach 0.8915, which is better than the traditional segmentation method.","PeriodicalId":373913,"journal":{"name":"Proceedings of the 2020 4th International Conference on Deep Learning Technologies","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116748779","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}
E-learning has changed the education style in the developed countries. However, in the developing nations such as the East African (EA) countries, the students are still challenged by the accessibility of online learning materials. In this paper, we sought to alleviate this issue by proposing a recommendation method that helps the students from the developing countries in selecting more appropriate e-learning resources. To achieve this goal, an e-learning dataset composes of 1237 students from three different universities in East Africa is used and the learners' information including contextual, demographic, and ratings predictions are hybridized by applying a developed knowledge-based computational model to generate the recommendations in a unified manner. Results from experimental evaluations are presented and discussed to demonstrate the benefits of the proposed system.
{"title":"A Hybrid Model for E-Learning Resources Recommendations in the Developing Countries","authors":"Jean-Pierre Niyigena, Qingshan Jiang","doi":"10.1145/3417188.3417211","DOIUrl":"https://doi.org/10.1145/3417188.3417211","url":null,"abstract":"E-learning has changed the education style in the developed countries. However, in the developing nations such as the East African (EA) countries, the students are still challenged by the accessibility of online learning materials. In this paper, we sought to alleviate this issue by proposing a recommendation method that helps the students from the developing countries in selecting more appropriate e-learning resources. To achieve this goal, an e-learning dataset composes of 1237 students from three different universities in East Africa is used and the learners' information including contextual, demographic, and ratings predictions are hybridized by applying a developed knowledge-based computational model to generate the recommendations in a unified manner. Results from experimental evaluations are presented and discussed to demonstrate the benefits of the proposed system.","PeriodicalId":373913,"journal":{"name":"Proceedings of the 2020 4th International Conference on Deep Learning Technologies","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116768694","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}
BIM (Building Information Modeling) technology is widely used in the field of traffic engineering. At this stage, more and more applied universities tries to improve the effect of visual teaching by BIM technology. It discussed the necessity of technological curriculum reform and pointed out the problems existing in the traditional teaching of road and bridge engineering courses. It introduced the method and effect of teaching reform of road and bridge engineering courses by creating the BIM visual teaching resource library, and presents the necessary conditions in this course reform, and finally put forward the idea of opening BIM technology application course for long-term development of road and bridge engineering specialty.
BIM (Building Information Modeling)技术在交通工程领域得到了广泛的应用。现阶段，越来越多的应用型高校尝试利用BIM技术提高可视化教学效果。论述了技术课程改革的必要性，指出了传统道路桥梁工程课程教学中存在的问题。介绍了通过创建BIM可视化教学资源库对道路桥梁工程课程进行教学改革的方法和效果，并提出了进行课程改革的必要条件，最后提出了为道路桥梁工程专业的长远发展开设BIM技术应用课程的设想。
{"title":"Research on the Innovation of BIM Technology in the Education of Road and Bridge Engineering Specialty","authors":"Man-li Tian, Ai-jun Jiang, Jie Wang","doi":"10.1145/3417188.3417207","DOIUrl":"https://doi.org/10.1145/3417188.3417207","url":null,"abstract":"BIM (Building Information Modeling) technology is widely used in the field of traffic engineering. At this stage, more and more applied universities tries to improve the effect of visual teaching by BIM technology. It discussed the necessity of technological curriculum reform and pointed out the problems existing in the traditional teaching of road and bridge engineering courses. It introduced the method and effect of teaching reform of road and bridge engineering courses by creating the BIM visual teaching resource library, and presents the necessary conditions in this course reform, and finally put forward the idea of opening BIM technology application course for long-term development of road and bridge engineering specialty.","PeriodicalId":373913,"journal":{"name":"Proceedings of the 2020 4th International Conference on Deep Learning Technologies","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125816479","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：