2022 International Conference on Intelligent Computing and Machine Learning (2ICML)最新文献

{"title":"lpData: A Data Placement for High-Throughput and Low-Latency","authors":"Huiying Zhang, Weixiang Zhang, Bo Wei, Qianran Si","doi":"10.1109/2ICML58251.2022.00012","DOIUrl":"https://doi.org/10.1109/2ICML58251.2022.00012","url":null,"abstract":"In the era of data explosion, many storage technologies have emerged for processing and analyzing the big data. Structured storage such as Parquet has highthroughput in sequential read while semi-structured storage such as HBase supports low-latency in random access. However, due to the gap between the two kinds of storage, neither of the storage is suitable to each other's application scenarios. Motivated by applications that need both access patterns, the work proposed a new data placement, that is, lpData. Inheriting efficient record-level index from Lucene and high-throughput file format from Parquet, lpData is able to speed up queries with predicates and guarantees the performance in sequential read at the same time. According to experimental results of this study, a) shows high performance in both sequential read and random access, b) compared to Parquet, lpData executes 42% faster on average in TPC-H selective queries, c) compared to Lucene, lpData outperforms 60% faster for low selective queries and by 36× for high selective queries on average.","PeriodicalId":355485,"journal":{"name":"2022 International Conference on Intelligent Computing and Machine Learning (2ICML)","volume":"399 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122864164","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":"NLP K-Means Algorithm Incorporated into a Proactive Chatbot to Assist Failing Students","authors":"Arlindo Almada, Qicheng Yu, Preeti Patel","doi":"10.1109/2ICML58251.2022.00015","DOIUrl":"https://doi.org/10.1109/2ICML58251.2022.00015","url":null,"abstract":"Predicting failure and individually assisting failing students is an ongoing challenge for most universities. This paper focuses on natural language processing and clustering the k-means algorithm applied to active chatbots. It aims to help students, and specifically to identify and predict failing students and proactively help them. Furthermore, it suggests an intervention to help students based on controllable academic factors that affect their academic performance. First, the authors outlined the research context for achieving this goal and created a predictive model of students' academic performance. The research results indicate a correlation between the variables with an accuracy of 0.935 and a precision of 0.76. Next, the k-means algorithm was used to cluster the students' problems or different factors that affect the students' academic performance. Finally, the k-means algorithm was integrated into an active chatbot to help students according to their problem groups.","PeriodicalId":355485,"journal":{"name":"2022 International Conference on Intelligent Computing and Machine Learning (2ICML)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123707813","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":"Exploration and Research of Waste Management System for High-Rise Buildings Based on Computer Technology","authors":"Caiying Huang, Ruiqing Shen, JingJing Wang, H. Cai","doi":"10.1109/2ICML58251.2022.00017","DOIUrl":"https://doi.org/10.1109/2ICML58251.2022.00017","url":null,"abstract":"The proper disposal of high-rise construction waste has become an important requirement for the standardization and efficiency of high-rise building construction. However, the current management process of high-rise buildings in China still faces a number of problems, including cluttered waste dumping sites, weak vertical transportation capacity, outbound conflicts leading to chaotic construction sites, and low construction efficiency. In order to solve these problems, the use of computer technology for intelligent management of construction waste has become a trend, and this paper relies on intelligent induction system and Internet of Things technology to establish an intelligent control platform for high-rise construction waste based on the use of computer technology. The system uses modular monitoring and removal tracking management to solve the problem of chaotic construction waste removal. In addition, this paper introduces a monitoring dust reduction system for dust detection and dust reduction treatment for construction waste removal, which is consistent with the green construction requirements of high-rise buildings. High-rise construction waste management system uses information technology as the main means to manage high-rise construction waste with information technology to overcome the drawbacks of traditional construction waste management and to effectively promote the construction process.","PeriodicalId":355485,"journal":{"name":"2022 International Conference on Intelligent Computing and Machine Learning (2ICML)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128396886","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":"Research on Fast Retrieval Algorithm for Sports Training Target Images","authors":"Jun Guo, Xiaoyu Wu","doi":"10.1109/2ICML58251.2022.00014","DOIUrl":"https://doi.org/10.1109/2ICML58251.2022.00014","url":null,"abstract":"Current traditional image retrieval methods achieve image retrieval by matching a single feature of an image, which leads to poor retrieval results due to the lack of pre-processing of the image. In this regard, the work proposes a study of a fast retrieval algorithm for sports training target images. By extracting visual features of sports training target images and generating corresponding visual vocabulary, denoising and visually enhancing the images, fast image retrieval is achieved by calculating image similarity measures. In the experiments, the proposed retrieval method is verified for retrieval performance. The experimental analysis shows that the proposed method has high retrieval speed and accuracy in retrieving images, and has a good overall retrieval performance.","PeriodicalId":355485,"journal":{"name":"2022 International Conference on Intelligent Computing and Machine Learning (2ICML)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121342182","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":"Research on the Transformer Fault Diagnosis Method Based on LSTM Artificial Neural Network and DGA","authors":"Z. Li, Yihua Qian, Qing Wang, Yaohong Zhao","doi":"10.1109/2ICML58251.2022.00018","DOIUrl":"https://doi.org/10.1109/2ICML58251.2022.00018","url":null,"abstract":"The concentration of dissolved gas in transformers is closely related to their operating status. Aiming at dissolved gas analysis (DGA) in transformer oil, this paper proposes a fault diagnosis method for transformer DGA based on long-term and short-term memory (LSTM) artificial neural networks. The method uses 240 sets of samples collected by China Southern Power Grid Corporation, with 180 sets as training data and the remaining 60 sets as test data. The input consists of five kinds of dissolved gases in oil, and the output is the corresponding fault type. The hyperparameters (H1=H2=50) of the network are determined through experimentation to establish a transformer DGA fault diagnosis model based on LSTM. The research results indicate that the LSTM diagnosis model has higher consistency with actual fault types compared to the traditional neural network diagnosis model. These findings demonstrate the promising application prospects of LSTM in the field of transformer DGA fault diagnosis.","PeriodicalId":355485,"journal":{"name":"2022 International Conference on Intelligent Computing and Machine Learning (2ICML)","volume":"102 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122417675","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":"Adaptive Segmentation of Basketball Game Video Based on Markov Random Fields","authors":"Lingfeng Yuan, Jing Shen, Ruisi Yang, Han Jiang","doi":"10.1109/2ICML58251.2022.00016","DOIUrl":"https://doi.org/10.1109/2ICML58251.2022.00016","url":null,"abstract":"To solve the problem of low coverage of some adaptive segmentation methods for basketball game videos, this paper proposes an adaptive segmentation method for basketball game videos based on Markov random field. Obtain the average value of pixel values of all frames at a specific position from the shot, extract the keyframes of the game video, and appropriately increase or reduce the segmentation threshold to resist illumination. The background model with texture information is constructed to obtain the smooth trajectory of basketball movement. The spatial position of each point in the video background is found in different frames. After obtaining a more accurate closed edge of the target, the pixel points are filled into it, and the adaptive segmentation process is optimized based on Markov random field. Experimental results show that the proposed adaptive segmentation method achieves an average coverage of 85.50% for basketball game videos, indicating its effectiveness after introducing Markov random field.","PeriodicalId":355485,"journal":{"name":"2022 International Conference on Intelligent Computing and Machine Learning (2ICML)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116922980","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":"Copyright Page","authors":"","doi":"10.1109/2icml58251.2022.00003","DOIUrl":"https://doi.org/10.1109/2icml58251.2022.00003","url":null,"abstract":"","PeriodicalId":355485,"journal":{"name":"2022 International Conference on Intelligent Computing and Machine Learning (2ICML)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130417542","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":"Recognition Algorithm of Basketball Dynamic Movement Behavior Based on Multimedia Network Technology","authors":"Xiaoyu Wu, Yumei Xue","doi":"10.1109/2ICML58251.2022.00020","DOIUrl":"https://doi.org/10.1109/2ICML58251.2022.00020","url":null,"abstract":"In order to realize the recognition of basketball's dynamic sports behavior in the dynamic scene of competition and training based on computer vision technology, combined with video data analysis, and improve the standardization of basketball's dynamic sports behavior, this paper puts forward a recognition method of basketball's dynamic sports behavior in the dynamic scene of competition and training based on multimedia network, and adopts the method of optical flow acquisition to collect RGB image data during the conversion process of basketball's dynamic sports behavior in the dynamic scene of competition and training. The collected optical flow field of basketball dynamic motion in the dynamic scene of competition and training is extracted with normative features, and the video parameters of basketball dynamic motion in the dynamic scene of competition and training are analyzed with the method of multimedia network technology video supervised learning. Under the constraint of dynamic basketball vision and optical flow field changing parameters, the temporal and spatial features extracted from basketball dynamic motion sequence are extracted by two-degree-of-freedom dynamic parameter model parameter learning. Combining the techniques of feature extraction, feature fusion and feature classification, this paper analyzes the optical flow trajectory parameters in the process of basketball dynamic movement behavior transformation, and realizes the dynamic parameter identification of basketball dynamic movement behavior in the dynamic scene of competition and training by combining the scene attributes and the distribution of time and space points of interest. The simulation results show that this method can effectively recognize the static and dynamic characteristics of basketball dynamic movement behavior in the dynamic scene of competition and training, and improve the dimension of basketball dynamic movement behavior recognition, thus effectively realizing the accurate detection and recognition of basketball dynamic movement behavior under complex background and multi-person interaction.","PeriodicalId":355485,"journal":{"name":"2022 International Conference on Intelligent Computing and Machine Learning (2ICML)","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127208215","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":"Deep Transfer Learning Optimization Techniques for Medical Image Classification: A Review","authors":"Paul Wahome Kariuki, P. Gikunda, J. Wandeto","doi":"10.1109/2ICML58251.2022.00013","DOIUrl":"https://doi.org/10.1109/2ICML58251.2022.00013","url":null,"abstract":"Medical image classification is a complex and challenging task due to the heterogeneous nature of medical data. Deep transfer learning has emerged as a promising technique for medical image classification, allowing the leveraging of knowledge from pre-trained models learned from large-scale datasets, resulting in improved performance with minimal training and overcoming the disadvantage of small data sets. This paper concisely overviews cutting-edge deep transfer learning optimization approaches for medical image classification. The study covers convolutional neural networks and transfer learning techniques, including relation-based, feature-based, parameter-based, and instance-based transfer learning. Classical classifiers such as Resnet, VGG, Alexnet, Googlenet, and Inception are examined, and their performance on medical image classification tasks is compared. The paper also discusses optimization techniques, such as batch normalization, regularization, and weight initialization, as well as data augmentation and kernel mathematical formulations. The study concludes by identifying challenges when using deep transfer learning for medical image classification and proposing potential future approaches for this field.","PeriodicalId":355485,"journal":{"name":"2022 International Conference on Intelligent Computing and Machine Learning (2ICML)","volume":"158 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121030681","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":"Specialized Risk Evaluation of Agricultural Products on Live-Streaming e-Commerce Platforms using Interval-Valued Intuitionistic Fuzzy Group Decision-Making","authors":"Dan Wang, Jiafu Su, Fengting Zhang","doi":"10.1109/2ICML58251.2022.00019","DOIUrl":"https://doi.org/10.1109/2ICML58251.2022.00019","url":null,"abstract":"Evaluating the risk of agricultural products live-streaming e-commerce platforms (abbreviation: platforms) is a key initiative for agricultural product merchants (abbreviation: merchants) to achieve sustainable development. For the uncertainty of the risk and risk evaluation environment and the fuzzy thinking of decision makers (DMs), this paper proposes an improved risk evaluation method based on interval-valued intuitionistic fuzzy (IVIF) multi-criteria group decision-making (MCGDM) to evaluate the risk of platforms. The method determines the DMs' weights for risk criteria based on the expertise of DMs in risk criteria and ranks the risk magnitude of alternative platforms using the VIse Kriterijumska Optimizacija I Kompromisno Resenje (VIKOR) method. The risk evaluation method proposed in this study is more suitable for the real situation of agricultural products live-streaming e-commerce platform. The research in this paper will not only provide practical guidance for merchants to determine the platforms with the lowest comprehensive risk, but also develop the theoretical research results in the field of agricultural products live-streaming e-commerce and help realize rural revitalization.","PeriodicalId":355485,"journal":{"name":"2022 International Conference on Intelligent Computing and Machine Learning (2ICML)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114724616","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}