Pub Date : 2021-12-01DOI: 10.1109/CSCI54926.2021.00147
Miku Kawai, Jumpei Ono, Takashi Ogata
This paper introduces the generation of explanations for a stage-performing structural simulation system using the animation of kabuki dance that we have developed. After considering the explanation function in the context of narratives and kabuki knowledge, we present a mechanism in the aforementioned system that automatically determines the content and method of an explanation according to the degree of the user’s interest and knowledge. Through this trial, we consider effective explanation methods for narrative generation and, in this study, the use of kabuki-related knowledge.
{"title":"Kabuki Explanation System Based on User’s Knowledge and Interests","authors":"Miku Kawai, Jumpei Ono, Takashi Ogata","doi":"10.1109/CSCI54926.2021.00147","DOIUrl":"https://doi.org/10.1109/CSCI54926.2021.00147","url":null,"abstract":"This paper introduces the generation of explanations for a stage-performing structural simulation system using the animation of kabuki dance that we have developed. After considering the explanation function in the context of narratives and kabuki knowledge, we present a mechanism in the aforementioned system that automatically determines the content and method of an explanation according to the degree of the user’s interest and knowledge. Through this trial, we consider effective explanation methods for narrative generation and, in this study, the use of kabuki-related knowledge.","PeriodicalId":206881,"journal":{"name":"2021 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"134 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134561180","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 : 2021-12-01DOI: 10.1109/CSCI54926.2021.00363
Marcel Danilescu, Laura Danilescu
Granting access to an organization's information resources is an issue that is the subject of numerous research works, with different approaches. This paper addresses access and action control policies based on the levels of trust given to them.The internal organization of an enterprise implies the existence of a hierarchy of departments, structured in a tree, in which data and information are circulated both ascending and descending. The staff of the organization is the one who carries out various processes, which consist of actions, workflows and/or information flows and events. In order to participate in these processes, a certain level of trust is assigned to the person. The association between the level of trust given to a person and the value of trust attributed to an object leads to the generation of policies implemented by computer applications that use access and actions control based on trust. The creation of these policies and their updating is done from a Policy Creation Point. A Policy Storage point is used to store all policies. The Document Status Point is the location where the document status matrix is located. Thus, The Document Storage Point is the space where documents are stored in electronic format. By creating them, a single point of access to policies is established for their creation and updating, a point where policies are stored, a point of storage of the workflow applied to documents and the active process, and a point of documents storage.Our paper presents, in addition to an original formal model, the use of trust gained by a member of an organization (trust calculated or attributed directly), and an example of its practical applicability in the information processes in the organization.This paper complements our previous work, which focuses on the aspects of using trust in controlling access and user interaction with information processes in the organization. This paper complements our previous work, which focuses on the aspects of using trust in controlling user access and interaction with information processes in the organization.
{"title":"Design of software applications using access and actions control policies based on trust","authors":"Marcel Danilescu, Laura Danilescu","doi":"10.1109/CSCI54926.2021.00363","DOIUrl":"https://doi.org/10.1109/CSCI54926.2021.00363","url":null,"abstract":"Granting access to an organization's information resources is an issue that is the subject of numerous research works, with different approaches. This paper addresses access and action control policies based on the levels of trust given to them.The internal organization of an enterprise implies the existence of a hierarchy of departments, structured in a tree, in which data and information are circulated both ascending and descending. The staff of the organization is the one who carries out various processes, which consist of actions, workflows and/or information flows and events. In order to participate in these processes, a certain level of trust is assigned to the person. The association between the level of trust given to a person and the value of trust attributed to an object leads to the generation of policies implemented by computer applications that use access and actions control based on trust. The creation of these policies and their updating is done from a Policy Creation Point. A Policy Storage point is used to store all policies. The Document Status Point is the location where the document status matrix is located. Thus, The Document Storage Point is the space where documents are stored in electronic format. By creating them, a single point of access to policies is established for their creation and updating, a point where policies are stored, a point of storage of the workflow applied to documents and the active process, and a point of documents storage.Our paper presents, in addition to an original formal model, the use of trust gained by a member of an organization (trust calculated or attributed directly), and an example of its practical applicability in the information processes in the organization.This paper complements our previous work, which focuses on the aspects of using trust in controlling access and user interaction with information processes in the organization. This paper complements our previous work, which focuses on the aspects of using trust in controlling user access and interaction with information processes in the organization.","PeriodicalId":206881,"journal":{"name":"2021 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130314875","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 : 2021-12-01DOI: 10.1109/CSCI54926.2021.00096
Korn Sooksatra, P. Rivas
With the increasing presence of deep learning models, many applications have had significant improvements; however, they face a new vulnerability known as adversarial examples. Adversarial examples can mislead deep learning models to predict the wrong classes without human actors noticing. Recently, many works have tried to improve adversarial examples to make them stronger and more effective. However, although some researchers have invented mechanisms to defend deep learning models against adversarial examples, those mechanisms may negatively affect different measures of fairness, which are critical in practice. This work mathematically defines four fairness scores to show that training adversarially robust models can harm fairness scores. Furthermore, we empirically show that adversarial training, one of the most potent defensive mechanisms against adversarial examples, can harm them.
{"title":"Adversarial Training Negatively Affects Fairness","authors":"Korn Sooksatra, P. Rivas","doi":"10.1109/CSCI54926.2021.00096","DOIUrl":"https://doi.org/10.1109/CSCI54926.2021.00096","url":null,"abstract":"With the increasing presence of deep learning models, many applications have had significant improvements; however, they face a new vulnerability known as adversarial examples. Adversarial examples can mislead deep learning models to predict the wrong classes without human actors noticing. Recently, many works have tried to improve adversarial examples to make them stronger and more effective. However, although some researchers have invented mechanisms to defend deep learning models against adversarial examples, those mechanisms may negatively affect different measures of fairness, which are critical in practice. This work mathematically defines four fairness scores to show that training adversarially robust models can harm fairness scores. Furthermore, we empirically show that adversarial training, one of the most potent defensive mechanisms against adversarial examples, can harm them.","PeriodicalId":206881,"journal":{"name":"2021 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130389019","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 : 2021-12-01DOI: 10.1109/CSCI54926.2021.00324
Lucas Mendes Lima, Victor Calebe Cavalcante, Mariana Guimarães de Sousa, Cláudio Afonso Fleury, D. Oliveira, Eduardo Noronha de Andrade Freitas
Context: Although agribusiness corresponded to more than 20% of Brazil’s Gross Domestic Product (GDP), most livestock is under manual control and manual monitoring. Additionally, alternative technologies are either uncomfortable and stressful, or expensive. Now, despite the great scientific advances in the area, there is still a pressing need for an automated robust, inexpensive and (sub)optimal technology to monitor animal behavior in a cost-effective, contact-less and stress-free fashion. Overall, this niche can leverage the benefits of Deep Learning schemes.Objective: This review aims to provide a systematic overview of most current projects in the area of comfort monitoring dairy cattle, as well as their corresponding image recognition-based techniques and technologies.Methods: First, a systematic review planning was carried out, and objectives, research questions, search strings, among others, were defined. Subsequently,a broad survey was conducted to extract, analyze and compile the data, to generate a easy-to-read visual source of information (tables and graphics).Results: Information was extracted from the reviewed papers. Among this data collected from the papers are techniques utilized, target behaviors, cow bodyparts identified in visual computational, besides their paper source font, the publication date, and localization. For example, the papers present are mostly recent. China has had a larger number of relevant papers in the area. The back was the body region most analyzed by the papers and the behaviors most analyzed were body condition score, lameness, cow’s body position and feeding/drinking behavior. Among the methods used is RCNN Inception V3 with the best accuracy for cow’s back region.Conclusion: The aim of this work is to present some of the papers that are being carried out in the area of dairy cow behavior monitoring, using techniques of Artifical Intelligence. It is expected that the information collected and presented in the present systematic review paper contribute to the future researches and projects of the area and the application of new techniques.
{"title":"Artificial Intelligence in Support of Welfare Monitoring of Dairy Cattle: A Systematic Literature Review","authors":"Lucas Mendes Lima, Victor Calebe Cavalcante, Mariana Guimarães de Sousa, Cláudio Afonso Fleury, D. Oliveira, Eduardo Noronha de Andrade Freitas","doi":"10.1109/CSCI54926.2021.00324","DOIUrl":"https://doi.org/10.1109/CSCI54926.2021.00324","url":null,"abstract":"Context: Although agribusiness corresponded to more than 20% of Brazil’s Gross Domestic Product (GDP), most livestock is under manual control and manual monitoring. Additionally, alternative technologies are either uncomfortable and stressful, or expensive. Now, despite the great scientific advances in the area, there is still a pressing need for an automated robust, inexpensive and (sub)optimal technology to monitor animal behavior in a cost-effective, contact-less and stress-free fashion. Overall, this niche can leverage the benefits of Deep Learning schemes.Objective: This review aims to provide a systematic overview of most current projects in the area of comfort monitoring dairy cattle, as well as their corresponding image recognition-based techniques and technologies.Methods: First, a systematic review planning was carried out, and objectives, research questions, search strings, among others, were defined. Subsequently,a broad survey was conducted to extract, analyze and compile the data, to generate a easy-to-read visual source of information (tables and graphics).Results: Information was extracted from the reviewed papers. Among this data collected from the papers are techniques utilized, target behaviors, cow bodyparts identified in visual computational, besides their paper source font, the publication date, and localization. For example, the papers present are mostly recent. China has had a larger number of relevant papers in the area. The back was the body region most analyzed by the papers and the behaviors most analyzed were body condition score, lameness, cow’s body position and feeding/drinking behavior. Among the methods used is RCNN Inception V3 with the best accuracy for cow’s back region.Conclusion: The aim of this work is to present some of the papers that are being carried out in the area of dairy cow behavior monitoring, using techniques of Artifical Intelligence. It is expected that the information collected and presented in the present systematic review paper contribute to the future researches and projects of the area and the application of new techniques.","PeriodicalId":206881,"journal":{"name":"2021 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115660002","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 : 2021-12-01DOI: 10.1109/CSCI54926.2021.00141
O. Kotevska, Folami T. Alamudun, Christopher Stanley
As the number of online services has increased, the amount of sensitive data being recorded is rising. Simultaneously, the decision-making process has improved by using the vast amounts of data, where machine learning has transformed entire industries. This paper addresses the development of optimal private deep neural networks and discusses the challenges associated with this task. We focus on differential privacy implementations and finding the optimal balance between accuracy and privacy, benefits and limitations of existing libraries, and challenges of applying private machine learning models in practical applications. Our analysis shows that learning rate, and privacy budget are the key factors that impact the results, and we discuss options for these settings.
{"title":"Optimal Balance of Privacy and Utility with Differential Privacy Deep Learning Frameworks","authors":"O. Kotevska, Folami T. Alamudun, Christopher Stanley","doi":"10.1109/CSCI54926.2021.00141","DOIUrl":"https://doi.org/10.1109/CSCI54926.2021.00141","url":null,"abstract":"As the number of online services has increased, the amount of sensitive data being recorded is rising. Simultaneously, the decision-making process has improved by using the vast amounts of data, where machine learning has transformed entire industries. This paper addresses the development of optimal private deep neural networks and discusses the challenges associated with this task. We focus on differential privacy implementations and finding the optimal balance between accuracy and privacy, benefits and limitations of existing libraries, and challenges of applying private machine learning models in practical applications. Our analysis shows that learning rate, and privacy budget are the key factors that impact the results, and we discuss options for these settings.","PeriodicalId":206881,"journal":{"name":"2021 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114418535","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 : 2021-12-01DOI: 10.1109/CSCI54926.2021.00259
Misagh Faezipour, M. Faezipour, Brianna Bauman
This paper proposes a systems engineering perspective to analyze the causes of COVID-19 health disparities impact and interventions to minimize the impact on minorities. The impact of the novel coronavirus has shown to be more intense on minorities. The percentage of COVID-19 case count and fatality rate for minorities is much higher than that of the general population, showing that they are more significantly affected than others. Many different factors influence this impact, ranging from economic to cultural. In this paper, these factors are shown to be connected through a causal model analyzing the effects of each factor, after which, potential interventions are suggested. Many factors are identified, such as high employment in the service industry or lower likelihood to have insurance. From this, a causal model is created showing the impact of each factor. Using this causal model, one can identify the high-impact factors causing a disparate impact as well as suggest possible interventions including making testing and treatment more accessible, reducing healthcare bias, and improving healthcare for immigrants.
{"title":"Development of a Causal Model to Study the Disparate Effects of COVID-19 on Minorities","authors":"Misagh Faezipour, M. Faezipour, Brianna Bauman","doi":"10.1109/CSCI54926.2021.00259","DOIUrl":"https://doi.org/10.1109/CSCI54926.2021.00259","url":null,"abstract":"This paper proposes a systems engineering perspective to analyze the causes of COVID-19 health disparities impact and interventions to minimize the impact on minorities. The impact of the novel coronavirus has shown to be more intense on minorities. The percentage of COVID-19 case count and fatality rate for minorities is much higher than that of the general population, showing that they are more significantly affected than others. Many different factors influence this impact, ranging from economic to cultural. In this paper, these factors are shown to be connected through a causal model analyzing the effects of each factor, after which, potential interventions are suggested. Many factors are identified, such as high employment in the service industry or lower likelihood to have insurance. From this, a causal model is created showing the impact of each factor. Using this causal model, one can identify the high-impact factors causing a disparate impact as well as suggest possible interventions including making testing and treatment more accessible, reducing healthcare bias, and improving healthcare for immigrants.","PeriodicalId":206881,"journal":{"name":"2021 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116745047","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 : 2021-12-01DOI: 10.1109/CSCI54926.2021.00137
Noah Oghenefego Ogwara, K. Petrova, M. Yang
This paper presents an ensemble machine learning (ML) based system for the detection of malicious applications in the Mobile Cloud Computing (MCC) Environment. The proposed system named MOBDroid2 applies a static feature analysis approach using the permissions and intents demanded by Android apps. The experiments conducted showed that the proposed system was able to effectively detect malicious and benign apps, achieving a classification accuracy rate of 98.16%, a precision rate of 98.95%, a recall rate of 98.20%, and a false alarm rate of 1.85%. The results obtained in our experiment compared well with other results reported in extant literature.
{"title":"MOBDroid2: An Improved Feature Selection Method for Detecting Malicious Applications in a Mobile Cloud Computing Environment","authors":"Noah Oghenefego Ogwara, K. Petrova, M. Yang","doi":"10.1109/CSCI54926.2021.00137","DOIUrl":"https://doi.org/10.1109/CSCI54926.2021.00137","url":null,"abstract":"This paper presents an ensemble machine learning (ML) based system for the detection of malicious applications in the Mobile Cloud Computing (MCC) Environment. The proposed system named MOBDroid2 applies a static feature analysis approach using the permissions and intents demanded by Android apps. The experiments conducted showed that the proposed system was able to effectively detect malicious and benign apps, achieving a classification accuracy rate of 98.16%, a precision rate of 98.95%, a recall rate of 98.20%, and a false alarm rate of 1.85%. The results obtained in our experiment compared well with other results reported in extant literature.","PeriodicalId":206881,"journal":{"name":"2021 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115396675","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 : 2021-12-01DOI: 10.1109/CSCI54926.2021.00149
Chuluunsukh Anudari, YoungSu Yun, M. Gen
A supply chain network (SCN) model which considers facility and route disruptions simultaneously is proposed in this paper. Since most of conventional literature have focused either on facility disruption solely or on route disruption solely, the simultaneous consideration of facility and route disruptions can improve the flexibility of the implementation in the SCN model. The SCN model under the disruptions is represented as a mathematical formulation and a hybrid meta-heuristics (GA-VNS) approach which combines genetic algorithm (GA) with variable neighborhood search (VNS) is used for the mathematical formulation. In numerical experiment, two scaled SCN models are used for comparing the performance of the GA-VNS approach with those of some conventional meta-heuristics approaches. Experimental results prove that the GA-VNS approach is more robust than conventional meta-heuristics approaches, and the flexibility of the SCN model under the disruptions are also improved.
{"title":"Hybrid Meta-heuristics Approach for Solving Supply Chain Network Model under Disruption Risk","authors":"Chuluunsukh Anudari, YoungSu Yun, M. Gen","doi":"10.1109/CSCI54926.2021.00149","DOIUrl":"https://doi.org/10.1109/CSCI54926.2021.00149","url":null,"abstract":"A supply chain network (SCN) model which considers facility and route disruptions simultaneously is proposed in this paper. Since most of conventional literature have focused either on facility disruption solely or on route disruption solely, the simultaneous consideration of facility and route disruptions can improve the flexibility of the implementation in the SCN model. The SCN model under the disruptions is represented as a mathematical formulation and a hybrid meta-heuristics (GA-VNS) approach which combines genetic algorithm (GA) with variable neighborhood search (VNS) is used for the mathematical formulation. In numerical experiment, two scaled SCN models are used for comparing the performance of the GA-VNS approach with those of some conventional meta-heuristics approaches. Experimental results prove that the GA-VNS approach is more robust than conventional meta-heuristics approaches, and the flexibility of the SCN model under the disruptions are also improved.","PeriodicalId":206881,"journal":{"name":"2021 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123162132","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 : 2021-12-01DOI: 10.1109/CSCI54926.2021.00232
D. Ahmed
Learning can be more efficient, effective and interesting if we can identify more about our students and know how they learn. Due to COVID-19, schools and colleges are offering online classes. It has a significant impact on students’ success. Therefore, course modality and teaching pedagogy need to be taken into consideration for crafting and creating instructional experiences that make leaning appealing and effective. A number of innovative teaching methods such as active learning, hybrid learning, social learning and flipped classrooms have been proposed and tested. Practically, several methods together can be helpful for students. In this study, I conducted an experiment and identified effective learning methods for graduate level courses. According to this study, 94% students responded positively about this course design. The results also show that 83.5% students prefer face-to-face classes and 97% students find in-class problem solving effective to understand a concept better. Many courses incorporate team-based learning which is a proven approach. In this study, the benefits and limitations of team-based programming projects are identified as well as students’ opinion in this regard. The results show that 85% students prefer team-based programming projects. Surprisingly 59.1% students mentioned all members do not contribute fairly evenly. This is a common problem in group works. So, small group size could be effective to overcome this problem.
{"title":"Discovering Effective Learning Methods and Impact of Team-based Programming Projects in Graduate Level Courses","authors":"D. Ahmed","doi":"10.1109/CSCI54926.2021.00232","DOIUrl":"https://doi.org/10.1109/CSCI54926.2021.00232","url":null,"abstract":"Learning can be more efficient, effective and interesting if we can identify more about our students and know how they learn. Due to COVID-19, schools and colleges are offering online classes. It has a significant impact on students’ success. Therefore, course modality and teaching pedagogy need to be taken into consideration for crafting and creating instructional experiences that make leaning appealing and effective. A number of innovative teaching methods such as active learning, hybrid learning, social learning and flipped classrooms have been proposed and tested. Practically, several methods together can be helpful for students. In this study, I conducted an experiment and identified effective learning methods for graduate level courses. According to this study, 94% students responded positively about this course design. The results also show that 83.5% students prefer face-to-face classes and 97% students find in-class problem solving effective to understand a concept better. Many courses incorporate team-based learning which is a proven approach. In this study, the benefits and limitations of team-based programming projects are identified as well as students’ opinion in this regard. The results show that 85% students prefer team-based programming projects. Surprisingly 59.1% students mentioned all members do not contribute fairly evenly. This is a common problem in group works. So, small group size could be effective to overcome this problem.","PeriodicalId":206881,"journal":{"name":"2021 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"13 24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124688574","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 : 2021-12-01DOI: 10.1109/CSCI54926.2021.00110
Julian Cagnazzo, Osama Sam Abuomar, A. Yanguas-Gil, J. Elam
Atomic layer deposition (ALD) is a chemical engineering process used to coat surfaces with a thin film. It is a versatile process able to deposit a wide range of films using different chemical reagents. When developing novel ALD processes, a technician must determine the dosing time of each reagent. To accelerate this development process, we trained convolutional neural networks to predict the reagent saturation times of novel ALD reactions given the reagent dosing times and film growth rates of example reactions. We generated two kinds of models. Single reaction models made predictions based on a single example ALD reaction. Multiple reaction models made predictions based on ten example reactions using the same reagents with different dosing times.
{"title":"Atomic Layer Deposition Optimization Using Convolutional Neural Networks","authors":"Julian Cagnazzo, Osama Sam Abuomar, A. Yanguas-Gil, J. Elam","doi":"10.1109/CSCI54926.2021.00110","DOIUrl":"https://doi.org/10.1109/CSCI54926.2021.00110","url":null,"abstract":"Atomic layer deposition (ALD) is a chemical engineering process used to coat surfaces with a thin film. It is a versatile process able to deposit a wide range of films using different chemical reagents. When developing novel ALD processes, a technician must determine the dosing time of each reagent. To accelerate this development process, we trained convolutional neural networks to predict the reagent saturation times of novel ALD reactions given the reagent dosing times and film growth rates of example reactions. We generated two kinds of models. Single reaction models made predictions based on a single example ALD reaction. Multiple reaction models made predictions based on ten example reactions using the same reagents with different dosing times.","PeriodicalId":206881,"journal":{"name":"2021 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"162 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121877953","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：