R. Mihail, Wesley I. Cook, Brandi M. Griffin, T. Uyeno, C. Anderson
Remote sensing has revolutionized the efficiency of vegetation mapping, but such techniques remain impractical for mapping some types of flora over relatively limited spatial extents. We propose a deep-learning based framework for automated detection and planar mapping of an epiphytic plant in a forest from geotagged static imagery using inexpensive cameras. Our pipeline consists of two steps: segmentation and spatial distribution estimation. We evaluate several segmentation methods on a novel dataset of roughly 375 outdoor images with per-pixel labels indicating the presence of Spanish moss. We also evaluate the accuracy of the spatial distribution estimates with respect to field measurements by ecologists for Spanish moss.
{"title":"Vegetation density estimation in the wild","authors":"R. Mihail, Wesley I. Cook, Brandi M. Griffin, T. Uyeno, C. Anderson","doi":"10.1145/3190645.3190690","DOIUrl":"https://doi.org/10.1145/3190645.3190690","url":null,"abstract":"Remote sensing has revolutionized the efficiency of vegetation mapping, but such techniques remain impractical for mapping some types of flora over relatively limited spatial extents. We propose a deep-learning based framework for automated detection and planar mapping of an epiphytic plant in a forest from geotagged static imagery using inexpensive cameras. Our pipeline consists of two steps: segmentation and spatial distribution estimation. We evaluate several segmentation methods on a novel dataset of roughly 375 outdoor images with per-pixel labels indicating the presence of Spanish moss. We also evaluate the accuracy of the spatial distribution estimates with respect to field measurements by ecologists for Spanish moss.","PeriodicalId":403177,"journal":{"name":"Proceedings of the ACMSE 2018 Conference","volume":"141 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123178539","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}
Previous research has been done that has explored the use of embodied conversational agents as mentors in faculty-student relationships. This previous system requires users to go to a unique website to interact with the conversational agent. This paper presents the design and development of a conversational agent mentor that uses a more pervasive application for dialogue, short message service (SMS). The SMS conversational agent is constructed to be used as a virtual mentor, to mentor undergraduate computer science majors at a Historically Black College (HBCU) who are considering pursuing a graduate degree in computing. A study has been designed to compare the effectiveness of the SMS conversational agent to the original conversational agent, an embodied conversational agent (ECA).
{"title":"Using SMS as an interface for a virtual mentoring system","authors":"Leron Julian, Kinnis Gosha","doi":"10.1145/3190645.3190709","DOIUrl":"https://doi.org/10.1145/3190645.3190709","url":null,"abstract":"Previous research has been done that has explored the use of embodied conversational agents as mentors in faculty-student relationships. This previous system requires users to go to a unique website to interact with the conversational agent. This paper presents the design and development of a conversational agent mentor that uses a more pervasive application for dialogue, short message service (SMS). The SMS conversational agent is constructed to be used as a virtual mentor, to mentor undergraduate computer science majors at a Historically Black College (HBCU) who are considering pursuing a graduate degree in computing. A study has been designed to compare the effectiveness of the SMS conversational agent to the original conversational agent, an embodied conversational agent (ECA).","PeriodicalId":403177,"journal":{"name":"Proceedings of the ACMSE 2018 Conference","volume":"114 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123116946","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}
The digital forensics tools, as well as the data collected by investigators, must ensure that integrity is kept throughout the investigation process, which is known as validation. Through this research project, an approach that validates mobile forensics tools and the data stored on the device will be developed. The goal of this research is to develop an approach that validates mobile forensics tools and the data stored on the device. An application will be implemented in a forensically sound manner using the approach, which will comprise of robust validation methods. This research will focus on iOS apps, whereas, the data collection is conducted on the iOS device and at the end of process, data is transferred to a laptop for validation. This tool is designed to assist in the validation of forensic evidence so that it holds admissibility in court and decrease the time spent acquiring data from a mobile device by leveraging an aimed approach.
{"title":"A framework for validating aimed mobile digital forensics evidences","authors":"Rodney Wilson, H. Chi","doi":"10.1145/3190645.3190695","DOIUrl":"https://doi.org/10.1145/3190645.3190695","url":null,"abstract":"The digital forensics tools, as well as the data collected by investigators, must ensure that integrity is kept throughout the investigation process, which is known as validation. Through this research project, an approach that validates mobile forensics tools and the data stored on the device will be developed. The goal of this research is to develop an approach that validates mobile forensics tools and the data stored on the device. An application will be implemented in a forensically sound manner using the approach, which will comprise of robust validation methods. This research will focus on iOS apps, whereas, the data collection is conducted on the iOS device and at the end of process, data is transferred to a laptop for validation. This tool is designed to assist in the validation of forensic evidence so that it holds admissibility in court and decrease the time spent acquiring data from a mobile device by leveraging an aimed approach.","PeriodicalId":403177,"journal":{"name":"Proceedings of the ACMSE 2018 Conference","volume":"616 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123268426","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}
T. Lux, L. Watson, Tyler H. Chang, Jon Bernard, Bo Li, Xiaodong Yu, Li Xu, Godmar Back, A. Butt, K. Cameron, D. Yao, Yili Hong
A rapid increase in the quantity of data available is allowing all fields of science to generate more accurate models of multivariate phenomena. Regression and interpolation become challenging when the dimension of data is large, especially while maintaining tractable computational complexity. This paper proposes three novel techniques for multivariate interpolation and regression that each have polynomial complexity with respect to number of instances (points) and number of attributes (dimension). Initial results suggest that these techniques are capable of effectively modeling multivariate phenomena while maintaining flexibility in different application domains.
{"title":"Novel meshes for multivariate interpolation and approximation","authors":"T. Lux, L. Watson, Tyler H. Chang, Jon Bernard, Bo Li, Xiaodong Yu, Li Xu, Godmar Back, A. Butt, K. Cameron, D. Yao, Yili Hong","doi":"10.1145/3190645.3190687","DOIUrl":"https://doi.org/10.1145/3190645.3190687","url":null,"abstract":"A rapid increase in the quantity of data available is allowing all fields of science to generate more accurate models of multivariate phenomena. Regression and interpolation become challenging when the dimension of data is large, especially while maintaining tractable computational complexity. This paper proposes three novel techniques for multivariate interpolation and regression that each have polynomial complexity with respect to number of instances (points) and number of attributes (dimension). Initial results suggest that these techniques are capable of effectively modeling multivariate phenomena while maintaining flexibility in different application domains.","PeriodicalId":403177,"journal":{"name":"Proceedings of the ACMSE 2018 Conference","volume":"79 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125120513","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}
The cloud computing systems face challenges from various aspects, including deployment cost, security, and performance. This project surveys these challenges and the approaches to addressing them. Deployment cost is an intense challenge to cloud computing. The high network bandwidth required to support potentially millions of users is a key aspect that leads to the rising costs for cloud service providers. To alleviate the bandwidth consumption, lightweight frameworks such as Angular have been developed to reduce the bandwidth necessary to transmit the required data by allowing clients to optimize and interpret the data rather than sending large files over a network. Video streaming codecs have adapted to keep high quality images, while lowering the amount of bandwidth needed to deliver the high definition video. Security is a critically important aspect of cloud services. User's data is at risk of being intercepted when it is transmitted across an unsecured network. Approaches to solving the problem of network security include data encryption and the use of Virtual Private Network (VPNs). Encryption prevents private data from leaking to unwanted users, with the cost of relatively high computational overhead. VPNs allow data to be delivered securely between authenticated end-points via a sub-network defined over a public underlying network. In addition to the networks, cloud servers themselves may be vulnerable to security attacks, both internally and externally. Internal attacks can be rogue administrators, exploiting cloud vulnerabilities, and using the cloud to conduct "nefarious activity". The damages range from stealing unprotected files to using the servers as tools to attack other organizations or users. External attacks include cross site scripting, information leakage and improper handling, broken authentication and session management, failure to restrict Web access, improper data validation, insecure communications, and malicious file execution. To resolve the security concerns on the cloud servers, various technologies have been developed, including storage with encryption, firewalls, packet filters, as well as virtual local area networks. Performance is another very important issue in cloud computing. Poor performance can be caused by various factors, including improper use of hardware resources on either end of the connection, and the network latency. From the hardware aspect, an effort to improve performance has been the move towards flash-based storage in order to increase read/write speeds and therefore decrease access times. In addition, parallel computing allows more work to be done in the same amount of time and significantly benefits large cloud server systems that run thousands of systems in parallel. Deploying small data centers closer to the end-users can help significantly reduce the latency perceived by users as well as the workload on the main datacenters, at the cost of increased deployment expense for cloud service pro
{"title":"Cloud computing: cost, security, and performance","authors":"Jeremiah Bolin, Mengkun Yang","doi":"10.1145/3190645.3190706","DOIUrl":"https://doi.org/10.1145/3190645.3190706","url":null,"abstract":"The cloud computing systems face challenges from various aspects, including deployment cost, security, and performance. This project surveys these challenges and the approaches to addressing them. Deployment cost is an intense challenge to cloud computing. The high network bandwidth required to support potentially millions of users is a key aspect that leads to the rising costs for cloud service providers. To alleviate the bandwidth consumption, lightweight frameworks such as Angular have been developed to reduce the bandwidth necessary to transmit the required data by allowing clients to optimize and interpret the data rather than sending large files over a network. Video streaming codecs have adapted to keep high quality images, while lowering the amount of bandwidth needed to deliver the high definition video. Security is a critically important aspect of cloud services. User's data is at risk of being intercepted when it is transmitted across an unsecured network. Approaches to solving the problem of network security include data encryption and the use of Virtual Private Network (VPNs). Encryption prevents private data from leaking to unwanted users, with the cost of relatively high computational overhead. VPNs allow data to be delivered securely between authenticated end-points via a sub-network defined over a public underlying network. In addition to the networks, cloud servers themselves may be vulnerable to security attacks, both internally and externally. Internal attacks can be rogue administrators, exploiting cloud vulnerabilities, and using the cloud to conduct \"nefarious activity\". The damages range from stealing unprotected files to using the servers as tools to attack other organizations or users. External attacks include cross site scripting, information leakage and improper handling, broken authentication and session management, failure to restrict Web access, improper data validation, insecure communications, and malicious file execution. To resolve the security concerns on the cloud servers, various technologies have been developed, including storage with encryption, firewalls, packet filters, as well as virtual local area networks. Performance is another very important issue in cloud computing. Poor performance can be caused by various factors, including improper use of hardware resources on either end of the connection, and the network latency. From the hardware aspect, an effort to improve performance has been the move towards flash-based storage in order to increase read/write speeds and therefore decrease access times. In addition, parallel computing allows more work to be done in the same amount of time and significantly benefits large cloud server systems that run thousands of systems in parallel. Deploying small data centers closer to the end-users can help significantly reduce the latency perceived by users as well as the workload on the main datacenters, at the cost of increased deployment expense for cloud service pro","PeriodicalId":403177,"journal":{"name":"Proceedings of the ACMSE 2018 Conference","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116492320","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 software development, the life cycle in general consists of four different phases, which includes the design phase, development phase, implementation phase, and testing phase. The phases are being followed to ensure production of high-quality software that meets the demands and expectation of its intended audience. Security development follows a similar life cycle known as the Security Development Life Cycle (SecDLC) [1] to keep improving security in the real world. SecDLC is comprised of four distinct stages: Assessment, Detection, Protection, and Response. The goal of the SecDLC is to maintain, preserve, monitor, and improve information security. This paper outlines the initial development of the hands-on labs that would address every stage of SecDLC and thus provide practical tools to educate cybersecurity professionals. The labs developed will be a part of a new Cybersecurity educational framework.
{"title":"Design of cybersecurity hands-on laboratory exercises using secDLC framework","authors":"H. Chi, Temilola Aderibigbe, E. Kalaimannan","doi":"10.1145/3190645.3190704","DOIUrl":"https://doi.org/10.1145/3190645.3190704","url":null,"abstract":"In software development, the life cycle in general consists of four different phases, which includes the design phase, development phase, implementation phase, and testing phase. The phases are being followed to ensure production of high-quality software that meets the demands and expectation of its intended audience. Security development follows a similar life cycle known as the Security Development Life Cycle (SecDLC) [1] to keep improving security in the real world. SecDLC is comprised of four distinct stages: Assessment, Detection, Protection, and Response. The goal of the SecDLC is to maintain, preserve, monitor, and improve information security. This paper outlines the initial development of the hands-on labs that would address every stage of SecDLC and thus provide practical tools to educate cybersecurity professionals. The labs developed will be a part of a new Cybersecurity educational framework.","PeriodicalId":403177,"journal":{"name":"Proceedings of the ACMSE 2018 Conference","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126298621","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}
This paper defines a model for a relative grading system that scales from low enrollment to large enrollment courses. The model allows peer grader to rank a small subset of submissions that are later merged into a global ranked set of all submissions for an assignment. This global set of ranked submissions and grades produced by expert graders(instructors and/or teachers assistants) who perform absolute grading of their own small subset of the submissions, are used to interpolate a cardinal grade for each submission of an assignment but excluding the ones graded by the experts. This model will be simulated to define the values of different parameters that produces valid and reliable grades. The research will also show that the model is scalable for different size courses by examining the effort of peer graders and expert graders.
{"title":"A model of a relative peer grading system for MOOCs","authors":"Timothy Holston, D. Wilkins","doi":"10.1145/3190645.3190684","DOIUrl":"https://doi.org/10.1145/3190645.3190684","url":null,"abstract":"This paper defines a model for a relative grading system that scales from low enrollment to large enrollment courses. The model allows peer grader to rank a small subset of submissions that are later merged into a global ranked set of all submissions for an assignment. This global set of ranked submissions and grades produced by expert graders(instructors and/or teachers assistants) who perform absolute grading of their own small subset of the submissions, are used to interpolate a cardinal grade for each submission of an assignment but excluding the ones graded by the experts. This model will be simulated to define the values of different parameters that produces valid and reliable grades. The research will also show that the model is scalable for different size courses by examining the effort of peer graders and expert graders.","PeriodicalId":403177,"journal":{"name":"Proceedings of the ACMSE 2018 Conference","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129153624","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}
Janelle C. Mason, K. Kyei, Darrion Long, Hannah Foster, William Nick, James Mayes, A. Esterline
This paper presents a computational framework for identity (initially about the culprit in a crime scene) based on Barwise's situation theory. Situations support information and can carry information about other situations. An utterance situation carries information about a described situation thanks to the constraints imposed by natural language. We are concerned with utterance situations (which we call an id-situation or id-case) in which identity judgments are made about the culprit in a crime scene, which is the corresponding described situation. The id-situation and crime scene along with various resource situations make up a case in the legal sense. Resource situations include such things as where a fingerprint was filed and where one took a facial image used to train a classifier. In an investigation, there may be several coordinated id-cases that use different means in search of the same judgment. We utilize Semantic Web standards to express cases in our framework. We have developed Web Ontology Language (OWL) ontologies to provide concepts and principled vocabularies for encoding our scenarios in Resource Description Framework (RDF), and we present an example of a SPARQL Protocol and RDF Query Language (SPARQL) query of one of our encodings that spans situations. To follow how evidence supports hypotheses on the identity of the culprit in a crime scene, we use Dempster-Shafer theory, which provides numerical values for evidence-based confidence one should have in our identities. We tightly integrate it with our ontologies by having the representation of a case per our ontologies present a network containing situations and stitched together by objects; evidence "flows" along this network, diminishing and combining. We review the modifications of Dempster-Shafer theory required when one goes from a closed world assumption, where the number of suspects or candidates is bounded, to an open world assumption, where there is an unbounded number of suspects or candidates. We review our plans regarding equational reasoning based on identities established in our id-cases, and we review the related issues regarding the meanings of Universal Resource Identifiers (URIs). URIs are the fundamental denoting expressions ("names") on the Semantic Web, and an account of identity using Semantic-Web resources must be clear on what is required for a URI to denote something. On the pre-Semantic Web, a URI just identifies the web-page its protocol accesses, but on the Semantic Web, URIs are used to identify all resources, causing confusion between a thing and a page about it. We review three positions on the meanings of denoting expression and note that they apply to URIs as used in our encodings of scenarios.
{"title":"A framework for identity based on situation theory","authors":"Janelle C. Mason, K. Kyei, Darrion Long, Hannah Foster, William Nick, James Mayes, A. Esterline","doi":"10.1145/3190645.3190713","DOIUrl":"https://doi.org/10.1145/3190645.3190713","url":null,"abstract":"This paper presents a computational framework for identity (initially about the culprit in a crime scene) based on Barwise's situation theory. Situations support information and can carry information about other situations. An utterance situation carries information about a described situation thanks to the constraints imposed by natural language. We are concerned with utterance situations (which we call an id-situation or id-case) in which identity judgments are made about the culprit in a crime scene, which is the corresponding described situation. The id-situation and crime scene along with various resource situations make up a case in the legal sense. Resource situations include such things as where a fingerprint was filed and where one took a facial image used to train a classifier. In an investigation, there may be several coordinated id-cases that use different means in search of the same judgment. We utilize Semantic Web standards to express cases in our framework. We have developed Web Ontology Language (OWL) ontologies to provide concepts and principled vocabularies for encoding our scenarios in Resource Description Framework (RDF), and we present an example of a SPARQL Protocol and RDF Query Language (SPARQL) query of one of our encodings that spans situations. To follow how evidence supports hypotheses on the identity of the culprit in a crime scene, we use Dempster-Shafer theory, which provides numerical values for evidence-based confidence one should have in our identities. We tightly integrate it with our ontologies by having the representation of a case per our ontologies present a network containing situations and stitched together by objects; evidence \"flows\" along this network, diminishing and combining. We review the modifications of Dempster-Shafer theory required when one goes from a closed world assumption, where the number of suspects or candidates is bounded, to an open world assumption, where there is an unbounded number of suspects or candidates. We review our plans regarding equational reasoning based on identities established in our id-cases, and we review the related issues regarding the meanings of Universal Resource Identifiers (URIs). URIs are the fundamental denoting expressions (\"names\") on the Semantic Web, and an account of identity using Semantic-Web resources must be clear on what is required for a URI to denote something. On the pre-Semantic Web, a URI just identifies the web-page its protocol accesses, but on the Semantic Web, URIs are used to identify all resources, causing confusion between a thing and a page about it. We review three positions on the meanings of denoting expression and note that they apply to URIs as used in our encodings of scenarios.","PeriodicalId":403177,"journal":{"name":"Proceedings of the ACMSE 2018 Conference","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131659957","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}
Grading a classic open-ended problem-solving type engineering test is very time consuming and labor intensive. In order to investigate test taker's extent of interpretation on the given problem statement and the ability of the employment of governing equations and solving skills, test takers have been asked to solve on papers by hand. Recently, we can easily find online learning systems providing variety of ways of automatic grading tools but they are still limited to multiple choice questions or at most short answer questions. Hence, any kinds of minor errors, such as typos, precision propagation and failure cascades, are still lack of consideration. In the study, an algorithm and sample grading tool programmed in HTML/PHP for simple and most frequently score-losing failure modes, typo and calculation uncertainty, are introduced. The generic solution procedure for the open-ended type question follows: 1) problem interpretation 2) isolation of unknowns in the chosen governing equation 3) plugging in knowns to obtain the final answer. The suggested approach is to divide aforementioned three steps into sub questions to the extent of each variable in the governing equation. The questions are sequentially related from researching the given properties to determining the ultimate target value. The sub question set can investigate the depth of the test taker's knowledge on the utilizing equation. That means whether the person comprehends the physical/chemical meaning of each term or each variable; the reason of any canceled or ignored term; and the solution process. Precision error or truncation error is generic in the engineering problem solving. Sequential sub questions are related in a way that such errors implied in the previous questions transfer to the consecutive questions. As the result, the error range enlarges with the number of sub questions. This phenomena is similar to that occurs due to the measurement uncertainty propagation in metrology More challenging and labor intensive part in grading is tracking failures that cascaded to the consecutive steps. Test grader has to recalculate the question from the beginning to the end plugging in both correct numbers and incorrect numbers. The suggested algorithm is to use Boolean at each sub question for each variable and each step among values: (1) typed in value at the corresponding step and the corresponding variable; (2) originally saved correct answer with set tolerance; and (3) newly estimated answer using incorrect answer typed in the previous step. The effect of the failures made in the previous steps as well as the steps where the incorrect answers were typed in are investigated. Referring to the result of such logical analysis, test takers will receive a detailed feedback on what kind of errors they made, like the feedback from classic human grading method. This method has been utilized for several practical exams and verified that the grading time can be enormously reduced as well as provi
{"title":"Grading algorithm for cascading failures and propagation of precision error for the engineering course exams","authors":"Dongjin Kim","doi":"10.1145/3190645.3190708","DOIUrl":"https://doi.org/10.1145/3190645.3190708","url":null,"abstract":"Grading a classic open-ended problem-solving type engineering test is very time consuming and labor intensive. In order to investigate test taker's extent of interpretation on the given problem statement and the ability of the employment of governing equations and solving skills, test takers have been asked to solve on papers by hand. Recently, we can easily find online learning systems providing variety of ways of automatic grading tools but they are still limited to multiple choice questions or at most short answer questions. Hence, any kinds of minor errors, such as typos, precision propagation and failure cascades, are still lack of consideration. In the study, an algorithm and sample grading tool programmed in HTML/PHP for simple and most frequently score-losing failure modes, typo and calculation uncertainty, are introduced. The generic solution procedure for the open-ended type question follows: 1) problem interpretation 2) isolation of unknowns in the chosen governing equation 3) plugging in knowns to obtain the final answer. The suggested approach is to divide aforementioned three steps into sub questions to the extent of each variable in the governing equation. The questions are sequentially related from researching the given properties to determining the ultimate target value. The sub question set can investigate the depth of the test taker's knowledge on the utilizing equation. That means whether the person comprehends the physical/chemical meaning of each term or each variable; the reason of any canceled or ignored term; and the solution process. Precision error or truncation error is generic in the engineering problem solving. Sequential sub questions are related in a way that such errors implied in the previous questions transfer to the consecutive questions. As the result, the error range enlarges with the number of sub questions. This phenomena is similar to that occurs due to the measurement uncertainty propagation in metrology More challenging and labor intensive part in grading is tracking failures that cascaded to the consecutive steps. Test grader has to recalculate the question from the beginning to the end plugging in both correct numbers and incorrect numbers. The suggested algorithm is to use Boolean at each sub question for each variable and each step among values: (1) typed in value at the corresponding step and the corresponding variable; (2) originally saved correct answer with set tolerance; and (3) newly estimated answer using incorrect answer typed in the previous step. The effect of the failures made in the previous steps as well as the steps where the incorrect answers were typed in are investigated. Referring to the result of such logical analysis, test takers will receive a detailed feedback on what kind of errors they made, like the feedback from classic human grading method. This method has been utilized for several practical exams and verified that the grading time can be enormously reduced as well as provi","PeriodicalId":403177,"journal":{"name":"Proceedings of the ACMSE 2018 Conference","volume":"77 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132586664","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}
Featurel selection is common in prediction tasks because it helps in reducing computation time as well as dimensionality of the data. A hybrid filter wrapper approach has been presented in this paper to detect network intrusion attacks using the genetic algorithm. The genetic algorithm is a popular search algorithm with wide applications in optimization problems like the TSP problem. One of the biggest advantages of the genetic algorithm is its continuous evolution towards better solutions. However, it does take a greedy approach, evaluating its strength against a fitness function, making it vulnerable to local optima. A certain amount of randomness at each generation can help us overcome this problem. In Network Intrusion Detection systems, the number of attacks is sometimes far less than the false alarm rate, causing the real attacks to be ignored. To overcome this problem, we propose an objective function which not only rewards higher score for higher accuracy, but also heavily penalizes false positives. Features are initially selected based on information gain and each feature is weighted differently based on domain knowledge, and then the selected subset of features is scored based on accuracy with higher penalty for false positives. In addition, crossover and mutation is carried out to allow for sufficient randomness in feature selection and avoid overfitting. Sample experimentation on the UNSW-NB15 dataset show that our approach performs much better compared to traditional methods and other state-of-the-art intrusion detection classification algorithms.
{"title":"A hybrid genetic algorithm for network intrusion detection","authors":"S. Bagui, Debarghya Nandi, S. Bagui","doi":"10.1145/3190645.3190702","DOIUrl":"https://doi.org/10.1145/3190645.3190702","url":null,"abstract":"Featurel selection is common in prediction tasks because it helps in reducing computation time as well as dimensionality of the data. A hybrid filter wrapper approach has been presented in this paper to detect network intrusion attacks using the genetic algorithm. The genetic algorithm is a popular search algorithm with wide applications in optimization problems like the TSP problem. One of the biggest advantages of the genetic algorithm is its continuous evolution towards better solutions. However, it does take a greedy approach, evaluating its strength against a fitness function, making it vulnerable to local optima. A certain amount of randomness at each generation can help us overcome this problem. In Network Intrusion Detection systems, the number of attacks is sometimes far less than the false alarm rate, causing the real attacks to be ignored. To overcome this problem, we propose an objective function which not only rewards higher score for higher accuracy, but also heavily penalizes false positives. Features are initially selected based on information gain and each feature is weighted differently based on domain knowledge, and then the selected subset of features is scored based on accuracy with higher penalty for false positives. In addition, crossover and mutation is carried out to allow for sufficient randomness in feature selection and avoid overfitting. Sample experimentation on the UNSW-NB15 dataset show that our approach performs much better compared to traditional methods and other state-of-the-art intrusion detection classification algorithms.","PeriodicalId":403177,"journal":{"name":"Proceedings of the ACMSE 2018 Conference","volume":"64 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124997885","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：