Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00181
Dragi Kimovski, C. Bauer, Narges Mehran, R.-C. Prodan
The Computing Continuum, covering Cloud, Fog, and Edge systems, promises to provide on-demand resource-as-a-service for Internet applications with diverse requirements, ranging from extremely low latency to high-performance processing. However, eminent challenges in automating the resources man-agement of Big Data pipelines across the Computing Continuum remain. The resource management and adaptation for Big Data pipelines across the Computing Continuum require significant research effort, as the current data processing pipelines are dynamic. In contrast, traditional resource management strategies are static, leading to inefficient pipeline scheduling and overly complex process deployment. To address these needs, we propose in this work a scheduling and adaptation approach implemented as a software tool to lower the technological barriers to the management of Big Data pipelines over the Computing Continuum. The approach separates the static scheduling from the run-time execution, em-powering domain experts with little infrastructure and software knowledge to take an active part in the Big Data pipeline adaptation. We conduct a feasibility study using a digital healthcare use case to validate our approach. We illustrate concrete scenarios supported by demonstrating how the scheduling and adaptation tool and its implementation automate the management of the lifecycle of a remote patient monitoring, treatment, and care pipeline.
{"title":"Big Data Pipeline Scheduling and Adaptation on the Computing Continuum","authors":"Dragi Kimovski, C. Bauer, Narges Mehran, R.-C. Prodan","doi":"10.1109/COMPSAC54236.2022.00181","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00181","url":null,"abstract":"The Computing Continuum, covering Cloud, Fog, and Edge systems, promises to provide on-demand resource-as-a-service for Internet applications with diverse requirements, ranging from extremely low latency to high-performance processing. However, eminent challenges in automating the resources man-agement of Big Data pipelines across the Computing Continuum remain. The resource management and adaptation for Big Data pipelines across the Computing Continuum require significant research effort, as the current data processing pipelines are dynamic. In contrast, traditional resource management strategies are static, leading to inefficient pipeline scheduling and overly complex process deployment. To address these needs, we propose in this work a scheduling and adaptation approach implemented as a software tool to lower the technological barriers to the management of Big Data pipelines over the Computing Continuum. The approach separates the static scheduling from the run-time execution, em-powering domain experts with little infrastructure and software knowledge to take an active part in the Big Data pipeline adaptation. We conduct a feasibility study using a digital healthcare use case to validate our approach. We illustrate concrete scenarios supported by demonstrating how the scheduling and adaptation tool and its implementation automate the management of the lifecycle of a remote patient monitoring, treatment, and care pipeline.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"2010 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127346773","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}
Distributed machine learning is essential for applying deep learning models with many data and parameters. Current researches on distributed machine learning focus on using more hardware devices powerful computing units for fast training. Consequently, the model training prefers a larger batch size to accelerate the training speed. However, the large batch training often suffers from poor accuracy due to poor generalization ability. Researchers have come up with many sophisticated methods to address this accuracy issue due to large batch sizes. These methods usually have complex mechanisms, thus making training more difficult. In addition, powerful training hardware for large batch sizes is expensive, and not all researchers can afford it. We propose a dual batch size learning scheme to address the batch size issue. We use the maximum batch size of our hardware for maximum training efficiency we can afford. In addition, we introduce a smaller batch size during the training to improve the model generalization ability. Using two different batch sizes in the same training simultaneously will reduce the testing loss and obtain a good generalization ability, with only a slight increase in the training time. We implement our dual batch size learning scheme and conduct experiments. By increasing 5% of the training time, we can reduce the loss from 1.429 to 1.246 in some cases. In addition, by appropriately adjusting the percentage of large and small batch sizes, we can increase the accuracy by 2.8% in some cases. With the additional 10% increase in training time, we can reduce the loss from 1.429 to 1.193. And after moderately adjusting the number of large batches and small batches used by GPUs, the accuracy can increase by 2.9%. Using two different batch sizes in the same training introduces two complications. First, the data processing speeds for two different batch sizes are different, so we must assign the data proportionally to maximize the overall processing speed. In addition, since the smaller batches will see fewer data due to the overall processing speed consideration, we proportionally adjust their contribution towards the global weight update in the parameter server. We use the ratio of data between the small and large batches to adjust the contribution. Experimental results indicate that this contribution adjustment increases the final accuracy by another 0.9%.
{"title":"Efficient Dual Batch Size Deep Learning for Distributed Parameter Server Systems","authors":"Kuan-Wei Lu, Pangfeng Liu, Ding-Yong Hong, Jan-Jan Wu","doi":"10.1109/COMPSAC54236.2022.00110","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00110","url":null,"abstract":"Distributed machine learning is essential for applying deep learning models with many data and parameters. Current researches on distributed machine learning focus on using more hardware devices powerful computing units for fast training. Consequently, the model training prefers a larger batch size to accelerate the training speed. However, the large batch training often suffers from poor accuracy due to poor generalization ability. Researchers have come up with many sophisticated methods to address this accuracy issue due to large batch sizes. These methods usually have complex mechanisms, thus making training more difficult. In addition, powerful training hardware for large batch sizes is expensive, and not all researchers can afford it. We propose a dual batch size learning scheme to address the batch size issue. We use the maximum batch size of our hardware for maximum training efficiency we can afford. In addition, we introduce a smaller batch size during the training to improve the model generalization ability. Using two different batch sizes in the same training simultaneously will reduce the testing loss and obtain a good generalization ability, with only a slight increase in the training time. We implement our dual batch size learning scheme and conduct experiments. By increasing 5% of the training time, we can reduce the loss from 1.429 to 1.246 in some cases. In addition, by appropriately adjusting the percentage of large and small batch sizes, we can increase the accuracy by 2.8% in some cases. With the additional 10% increase in training time, we can reduce the loss from 1.429 to 1.193. And after moderately adjusting the number of large batches and small batches used by GPUs, the accuracy can increase by 2.9%. Using two different batch sizes in the same training introduces two complications. First, the data processing speeds for two different batch sizes are different, so we must assign the data proportionally to maximize the overall processing speed. In addition, since the smaller batches will see fewer data due to the overall processing speed consideration, we proportionally adjust their contribution towards the global weight update in the parameter server. We use the ratio of data between the small and large batches to adjust the contribution. Experimental results indicate that this contribution adjustment increases the final accuracy by another 0.9%.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124924676","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}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00176
K. Lavania, Gaurang Gupta, D. Kumar
Documents are a common method of storing infor-mation and one of the most conventional forms of expression of ideas. Cloud servers store a user's documents with thousands of other users in place of physical storage devices. Indexes corresponding to the documents are also stored at the cloud server to enable the users to retrieve documents of their interest. The index includes keywords, document identities in which the keywords appear, along with Term Frequency-Inverse Document Frequency (TF-IDF) values which reflect the keywords' relevance scores of the dataset. Currently, there are no efficient methods to delete keywords from millions of documents over cloud servers while avoiding any compromise to the user's privacy. Most of the existing approaches use algorithms that divide a bigger problem into sub-problems and then combine them like divide and conquer problems. These approaches don't focus entirely on fine-grained deletion. This work is focused on achieving fine-grained deletion of keywords by keeping the size of the TF-IDF matrix constant after processing the deletion query, which comprises of keywords to be deleted. The experimental results of the proposed approach confirm that the precision of ranked search still remains very high after deletion without recalculation of the TF-IDF matrix.
{"title":"A Secure and Efficient Fine-Grained Deletion Approach over Encrypted Data","authors":"K. Lavania, Gaurang Gupta, D. Kumar","doi":"10.1109/COMPSAC54236.2022.00176","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00176","url":null,"abstract":"Documents are a common method of storing infor-mation and one of the most conventional forms of expression of ideas. Cloud servers store a user's documents with thousands of other users in place of physical storage devices. Indexes corresponding to the documents are also stored at the cloud server to enable the users to retrieve documents of their interest. The index includes keywords, document identities in which the keywords appear, along with Term Frequency-Inverse Document Frequency (TF-IDF) values which reflect the keywords' relevance scores of the dataset. Currently, there are no efficient methods to delete keywords from millions of documents over cloud servers while avoiding any compromise to the user's privacy. Most of the existing approaches use algorithms that divide a bigger problem into sub-problems and then combine them like divide and conquer problems. These approaches don't focus entirely on fine-grained deletion. This work is focused on achieving fine-grained deletion of keywords by keeping the size of the TF-IDF matrix constant after processing the deletion query, which comprises of keywords to be deleted. The experimental results of the proposed approach confirm that the precision of ranked search still remains very high after deletion without recalculation of the TF-IDF matrix.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126634058","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}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00023
Nearchos Paspallis, Nicos Kasenides, Andriani Piki
This paper presents an architecture for building multi player games that aim to teach coding skills and promote algorithmic thinking. The main requirements for the architecture are to enable quick and affordable development and deployment, support commodity client devices, and enable multiplayer, com-petitive gameplay. By demonstrating an evaluation case study, we show how the proposed architecture achieves these requirements. At its core, it realizes a distributed model extending the client-server paradigm, where multiple players can independently train, then compete in a multiplayer mode using a shared, cloud-based server. While the architecture is validated with a specific maze-themed case study game, we argue that the main principles of this approach can be reused to a wider range of multi player, educational games.
{"title":"A Software Architecture for Developing Distributed Games that Teach Coding and Algorithmic Thinking","authors":"Nearchos Paspallis, Nicos Kasenides, Andriani Piki","doi":"10.1109/COMPSAC54236.2022.00023","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00023","url":null,"abstract":"This paper presents an architecture for building multi player games that aim to teach coding skills and promote algorithmic thinking. The main requirements for the architecture are to enable quick and affordable development and deployment, support commodity client devices, and enable multiplayer, com-petitive gameplay. By demonstrating an evaluation case study, we show how the proposed architecture achieves these requirements. At its core, it realizes a distributed model extending the client-server paradigm, where multiple players can independently train, then compete in a multiplayer mode using a shared, cloud-based server. While the architecture is validated with a specific maze-themed case study game, we argue that the main principles of this approach can be reused to a wider range of multi player, educational games.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115102101","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}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00160
Motoyuki Ohmori
A L1 and/or L2 loop in a network may cause congestion and incur communication failures. It is, therefore, important to quickly and accurately detect a loop and locate its origin in order to eliminate the loop. To address this issue, this paper proposes a novel passive loop detection on edge ports of edge switches where end users' routers or terminals are accom-modated. The basic idea of the proposed detection is inspired by the nature that a MAC address of a L3 core switch should never be observed on an edge port. The MAC address is then observed by always-accepting MAC address authentication that can be easily deployed. The proposed detection can, therefore, accurately locate an edge port where a loop is formed, and avoid a failure to notify a network operator of a loop. In addition, the proposed detection can reduce a load on an edge switch more than the existing active detecting methods. Our evaluations in the real campus network have shown that the proposed method can detect the loop even where the existing methods cannot.
{"title":"A Novel Passive L1/L2 Edge Loop Detection Observing MAC addresses of L3 Core Switches","authors":"Motoyuki Ohmori","doi":"10.1109/COMPSAC54236.2022.00160","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00160","url":null,"abstract":"A L1 and/or L2 loop in a network may cause congestion and incur communication failures. It is, therefore, important to quickly and accurately detect a loop and locate its origin in order to eliminate the loop. To address this issue, this paper proposes a novel passive loop detection on edge ports of edge switches where end users' routers or terminals are accom-modated. The basic idea of the proposed detection is inspired by the nature that a MAC address of a L3 core switch should never be observed on an edge port. The MAC address is then observed by always-accepting MAC address authentication that can be easily deployed. The proposed detection can, therefore, accurately locate an edge port where a loop is formed, and avoid a failure to notify a network operator of a loop. In addition, the proposed detection can reduce a load on an edge switch more than the existing active detecting methods. Our evaluations in the real campus network have shown that the proposed method can detect the loop even where the existing methods cannot.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116096030","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}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00243
Zhenyu Liu, Xuanyu Lou, Yajun Cui, Ying Zhao, Hua Li
With the development of software defined network and network function virtualization, network operators can flexibly deploy service function chains (SFC) to provide network security services more than before according to the network security requirements of business systems. At present, most research on verifying the correctness of SFC is based on whether the logical sequence between service functions (SF) in SFC is correct before deployment, and there is less research on verifying the correctness after SFC deployment. Therefore, this paper proposes a method of using Colored Petri Net (CPN) to establish a verification model offline and verify whether each SF deployment in SFC is correct after online deployment. After the SFC deployment is completed, the information is obtained online and input into the established model for verification. The experimental results show that the SFC correctness verification method proposed in this paper can effectively verify whether each SF in the deployed SFC is deployed correctly. In this process, the correctness of SF model is verified by using SF model in the model library, and the model reuse technology is preliminarily discussed.
随着软件定义网络和网络功能虚拟化的发展，网络运营商可以根据业务系统的网络安全需求，灵活部署业务功能链(SFC)，提供比以往更多的网络安全服务。目前，验证SFC正确性的研究大多是基于部署前SFC中业务功能之间的逻辑顺序是否正确，部署后验证SFC正确性的研究较少。因此，本文提出了一种使用彩色Petri网(Colored Petri Net, CPN)离线建立验证模型的方法，在线部署后验证SFC中每个SF的部署是否正确。SFC部署完成后，在线获取信息并输入到建立的模型中进行验证。实验结果表明，本文提出的SFC正确性验证方法可以有效验证所部署SFC中的每个SF是否正确部署。在此过程中，通过在模型库中使用SF模型验证了SF模型的正确性，并对模型重用技术进行了初步探讨。
{"title":"Colored Petri Net Reusing for Service Function Chaining Validation","authors":"Zhenyu Liu, Xuanyu Lou, Yajun Cui, Ying Zhao, Hua Li","doi":"10.1109/COMPSAC54236.2022.00243","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00243","url":null,"abstract":"With the development of software defined network and network function virtualization, network operators can flexibly deploy service function chains (SFC) to provide network security services more than before according to the network security requirements of business systems. At present, most research on verifying the correctness of SFC is based on whether the logical sequence between service functions (SF) in SFC is correct before deployment, and there is less research on verifying the correctness after SFC deployment. Therefore, this paper proposes a method of using Colored Petri Net (CPN) to establish a verification model offline and verify whether each SF deployment in SFC is correct after online deployment. After the SFC deployment is completed, the information is obtained online and input into the established model for verification. The experimental results show that the SFC correctness verification method proposed in this paper can effectively verify whether each SF in the deployed SFC is deployed correctly. In this process, the correctness of SF model is verified by using SF model in the model library, and the model reuse technology is preliminarily discussed.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122670481","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}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00032
E. Corino, C. Fissore, M. Marchisio
In teaching it has become increasingly important to use didactic approaches that see students active and protagonists of their own learning. These approaches can often be supported by technologies, which also enable students to acquire digital skills and provide them with immediate and interactive feedback. In this paper we present recent research activities characterized by Data Driven Learning methodologies within a Digital Learning Environment integrated with an automatic formative assessment system to propose activities on the specialized language of Mathematics. In fact, Mathematics has always been one of the school disciplines in which students of all grades encounter the greatest difficulties. Numerous studies in Mathematics education have shown that the causes of disciplinary learning difficulties are the acquisition, understanding and management of one's language for specific purposes. The research activity involved 4 classes of two Italian secondary schools for a total of 80 students of grade 11 and their teachers. In this paper we study the impact that this type of activity has had on students, analyzing the students' responses to the final satisfaction questionnaire.
{"title":"Data Driven Learning activities within a Digital Learning Environment to study the specialized language of Mathematics","authors":"E. Corino, C. Fissore, M. Marchisio","doi":"10.1109/COMPSAC54236.2022.00032","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00032","url":null,"abstract":"In teaching it has become increasingly important to use didactic approaches that see students active and protagonists of their own learning. These approaches can often be supported by technologies, which also enable students to acquire digital skills and provide them with immediate and interactive feedback. In this paper we present recent research activities characterized by Data Driven Learning methodologies within a Digital Learning Environment integrated with an automatic formative assessment system to propose activities on the specialized language of Mathematics. In fact, Mathematics has always been one of the school disciplines in which students of all grades encounter the greatest difficulties. Numerous studies in Mathematics education have shown that the causes of disciplinary learning difficulties are the acquisition, understanding and management of one's language for specific purposes. The research activity involved 4 classes of two Italian secondary schools for a total of 80 students of grade 11 and their teachers. In this paper we study the impact that this type of activity has had on students, analyzing the students' responses to the final satisfaction questionnaire.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122100485","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}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00076
S. Reddivari
Requirements engineering (RE) is a crucial component in successful software development process. The idea of embedding the non-functional requirements (NFRs) such as performance, maintainability, modifiability, and others into a new software system is often implemented by software engineers. However, security as a crucial NFR is often ignored in the software development process. In this paper we address the importance of security as a NFR in the software development process. To that end, we propose a lightweight novel agile framework to analyze security requirements. We evaluate the proposed framework with a qualitative analysis and determine how it is useful to requirement analysts.
{"title":"An Agile Framework for Security Requirements: A Preliminary Investigation","authors":"S. Reddivari","doi":"10.1109/COMPSAC54236.2022.00076","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00076","url":null,"abstract":"Requirements engineering (RE) is a crucial component in successful software development process. The idea of embedding the non-functional requirements (NFRs) such as performance, maintainability, modifiability, and others into a new software system is often implemented by software engineers. However, security as a crucial NFR is often ignored in the software development process. In this paper we address the importance of security as a NFR in the software development process. To that end, we propose a lightweight novel agile framework to analyze security requirements. We evaluate the proposed framework with a qualitative analysis and determine how it is useful to requirement analysts.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128376500","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}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00074
Dan Lo, Kai Qian, Yong Shi, H. Shahriar, Chung Ng
Data science, machine learning, and distributed computational models have evolved dramatically over the last decade. Cloud and cluster computing is full-fledged and ready for processing big data. Data driven research and decision have become the trend in multiple disciplines. However, very few organizations have experienced the full impact or competitive advantage from their advanced data analytics initiatives despite significant investments in data science and machine learning. There are a number of issues resulting in such a phenomenon including difficult to maintain and configure a cluster, complex transition from a platform to another, sophisticated programming interfaces to machine learning libraries, network congestion, and most importantly lake of well-trained personnel to sanitize and analyze data. We propose a flexible heterogeneous computing cluster with off-the-shelf computers and a Blockly programming interface for multidisciplinary users such as cybersecurity ana-lyst, biologist, geologist, musician, and choreographer.
{"title":"Broaden Multidisciplinary Data Science Research by an Innovative Cyberinfrastructure Platform","authors":"Dan Lo, Kai Qian, Yong Shi, H. Shahriar, Chung Ng","doi":"10.1109/COMPSAC54236.2022.00074","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00074","url":null,"abstract":"Data science, machine learning, and distributed computational models have evolved dramatically over the last decade. Cloud and cluster computing is full-fledged and ready for processing big data. Data driven research and decision have become the trend in multiple disciplines. However, very few organizations have experienced the full impact or competitive advantage from their advanced data analytics initiatives despite significant investments in data science and machine learning. There are a number of issues resulting in such a phenomenon including difficult to maintain and configure a cluster, complex transition from a platform to another, sophisticated programming interfaces to machine learning libraries, network congestion, and most importantly lake of well-trained personnel to sanitize and analyze data. We propose a flexible heterogeneous computing cluster with off-the-shelf computers and a Blockly programming interface for multidisciplinary users such as cybersecurity ana-lyst, biologist, geologist, musician, and choreographer.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"92 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128643905","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}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00055
Junwei Lv, Xuegang Hu
Early classification of time series aims to accurately predict the class label of a time series as early as possible, which is significant but challenging in many time-sensitive applications. Existing early classification methods hold a basic closed-world assumption that the classifier must have seen the classes of test samples. However, new samples that do not belong to any trained class may appear in the real world. In this paper, we first address the early classification in an open world and design two detectors to identify which known class or unknown class a sample belongs to. Specifically, based on the observed data, an early known-class detector is designed to determine the known-class confidence and an early unknown-class detector is designed to determine the unknown-class confidence according to the Minimum Reliable Length (MRL) and the Weibull distribution of each class. Experimental results evaluated on real-world datasets demonstrate that the proposed model can identify samples of unknown and known classes accurately and early.
{"title":"A Baseline for Early Classification of Time Series in An Open World","authors":"Junwei Lv, Xuegang Hu","doi":"10.1109/COMPSAC54236.2022.00055","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00055","url":null,"abstract":"Early classification of time series aims to accurately predict the class label of a time series as early as possible, which is significant but challenging in many time-sensitive applications. Existing early classification methods hold a basic closed-world assumption that the classifier must have seen the classes of test samples. However, new samples that do not belong to any trained class may appear in the real world. In this paper, we first address the early classification in an open world and design two detectors to identify which known class or unknown class a sample belongs to. Specifically, based on the observed data, an early known-class detector is designed to determine the known-class confidence and an early unknown-class detector is designed to determine the unknown-class confidence according to the Minimum Reliable Length (MRL) and the Weibull distribution of each class. Experimental results evaluated on real-world datasets demonstrate that the proposed model can identify samples of unknown and known classes accurately and early.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129453948","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：