The support vector machine （SVM） is a novel machine learning method, which has the ability to approximate nonlinear functions with arbitrary accuracy. Setting parameters well is very crucial for SVM learning results...The support vector machine （SVM） is a novel machine learning method, which has the ability to approximate nonlinear functions with arbitrary accuracy. Setting parameters well is very crucial for SVM learning results and generalization ability, and now there is no systematic, general method for parameter selection. In this article, the SVM parameter selection for function approximation is regarded as a compound optimization problem and a mutative scale chaos optimization algorithm is employed to search for optimal paraxneter values. The chaos optimization algorithm is an effective way for global optimal and the mutative scale chaos algorithm could improve the search efficiency and accuracy. Several simulation examples show the sensitivity of the SVM parameters and demonstrate the superiority of this proposed method for nonlinear function approximation.展开更多
Research on the role of sexual selection in the speciation process largely focuses on the diversifying role of mate choice. In particular, much attention has been drawn to the fact that population divergence in mate c...Research on the role of sexual selection in the speciation process largely focuses on the diversifying role of mate choice. In particular, much attention has been drawn to the fact that population divergence in mate choice and in the male traits subject to choice directly can lead to assortative mating. However, male contest competition over mates also constitutes an important mechanism of sexual selection. We review recent empirical studies and argue that sexual selection through male contest competition can affect speciation in ways other than mate choice. For example, biases in aggression towards similar competitors can lead to disruptive and negative frequency-dependent selection on the traits used in contest competition in a similar way as competition for other types of limited resources. Moreover, male contest abilities often trade-off against other abilities such as parasite resistance, protection against predators and general stress tolerance. Populations experiencing different ecological condi- tions should therefore quickly diverge non-randomly in a number of traits including male contest abilities. In resource based breeding systems, a feedback loop between competitive ability and habitat use may lead to further population divergence. We discuss how population divergence in traits used in male contest competition can lead to the build up of reproductive isolation through a number of different pathways. Our main conclusion is that the role of male contest competition in speciation remains largely scientifically unexplored [Current Zoology 58 （3）： 493-509, 2012].展开更多
The autonomous navigation of an Unmanned Aerial Vehicle(UAV)relies heavily on the navigation sensors.The UAV’s level of autonomy depends upon the various navigation systems,such as state measurement,mapping,and obsta...The autonomous navigation of an Unmanned Aerial Vehicle(UAV)relies heavily on the navigation sensors.The UAV’s level of autonomy depends upon the various navigation systems,such as state measurement,mapping,and obstacle avoidance.Selecting the correct components is a critical part of the design process.However,this can be a particularly difficult task,especially for novices as there are several technologies and components available on the market,each with their own individual advantages and disadvantages.For example,satellite-based navigation components should be avoided when designing indoor UAVs.Incorporating them in the design brings no added value to the final product and will simply lead to increased cost and power consumption.Another issue is the number of vendors on the market,each trying to sell their hardware solutions which often incorporate similar technologies.The aim of this paper is to serve as a guide,proposing various methods to support the selection of fit-for-purpose technologies and components whilst avoiding system layout conflicts.The paper presents a study of the various navigation technologies and supports engineers in the selection of specific hardware solutions based on given requirements.The selection methods are based on easy-to-follow flow charts.A comparison of the various hardware components specifications is also included as part of this work.展开更多
Currently,the Internet of Things(IoT)is revolutionizing communi-cation technology by facilitating the sharing of information between different physical devices connected to a network.To improve control,customization,f...Currently,the Internet of Things(IoT)is revolutionizing communi-cation technology by facilitating the sharing of information between different physical devices connected to a network.To improve control,customization,flexibility,and reduce network maintenance costs,a new Software-Defined Network(SDN)technology must be used in this infrastructure.Despite the various advantages of combining SDN and IoT,this environment is more vulnerable to various attacks due to the centralization of control.Most methods to ensure IoT security are designed to detect Distributed Denial-of-Service(DDoS)attacks,but they often lack mechanisms to mitigate their severity.This paper proposes a Multi-Attack Intrusion Detection System(MAIDS)for Software-Defined IoT Networks(SDN-IoT).The proposed scheme uses two machine-learning algorithms to improve detection efficiency and provide a mechanism to prevent false alarms.First,a comparative analysis of the most commonly used machine-learning algorithms to secure the SDN was performed on two datasets:the Network Security Laboratory Knowledge Discovery in Databases(NSL-KDD)and the Canadian Institute for Cyberse-curity Intrusion Detection Systems(CICIDS2017),to select the most suitable algorithms for the proposed scheme and for securing SDN-IoT systems.The algorithms evaluated include Extreme Gradient Boosting(XGBoost),K-Nearest Neighbor(KNN),Random Forest(RF),Support Vector Machine(SVM),and Logistic Regression(LR).Second,an algorithm for selecting the best dataset for machine learning in Intrusion Detection Systems(IDS)was developed to enable effective comparison between the datasets used in the development of the security scheme.The results showed that XGBoost and RF are the best algorithms to ensure the security of SDN-IoT and to be applied in the proposed security system,with average accuracies of 99.88%and 99.89%,respectively.Furthermore,the proposed security scheme reduced the false alarm rate by 33.23%,which is a significant improvement over prevalent schemes.Finally,tests of the algorithm 展开更多
Traditional security systems are exposed to many various attacks,which represents a major challenge for the spread of the Internet in the future.Innovative techniques have been suggested for detecting attacks using ma...Traditional security systems are exposed to many various attacks,which represents a major challenge for the spread of the Internet in the future.Innovative techniques have been suggested for detecting attacks using machine learning and deep learning.The significant advantage of deep learning is that it is highly efficient,but it needs a large training time with a lot of data.Therefore,in this paper,we present a new feature reduction strategy based on Distributed Cumulative Histograms(DCH)to distinguish between dataset features to locate the most effective features.Cumulative histograms assess the dataset instance patterns of the applied features to identify the most effective attributes that can significantly impact the classification results.Three different models for detecting attacks using Convolutional Neural Network(CNN)and Long Short-Term Memory Network(LSTM)are also proposed.The accuracy test of attack detection using the hybrid model was 98.96%on the UNSW-NP15 dataset.The proposed model is compared with wrapper-based and filter-based Feature Selection(FS)models.The proposed model reduced classification time and increased detection accuracy.展开更多
Acoustic quality detection is vital in the manufactured products quality control field since it represents the conditions of machines or products.Recent work employed machine learning models in manufactured audio dat...Acoustic quality detection is vital in the manufactured products quality control field since it represents the conditions of machines or products.Recent work employed machine learning models in manufactured audio data to detect anomalous patterns.A major challenge is how to select applicable audio features to meliorate model’s accuracy and precision.To relax this challenge,we extract and analyze three audio feature types including Time Domain Feature,Frequency Domain Feature,and Cepstrum Feature to help identify the potential linear and non-linear relationships.In addition,we design a visual analysis system,namely AFExplorer,to assist data scientists in extracting audio features and selecting potential feature combinations.AFExplorer integrates four main views to present detailed distribution and relevance of the audio features,which helps users observe the impact of features visually in the feature selection.We perform the case study with AFExplore according to the ToyADMOS and MIMII Dataset to demonstrate the usability and effectiveness of the proposed system.展开更多
The development of new technologically advanced products requires the contribution from a range of skills and disciplines, which are often difficult to fred within a single company or organization. Requirements establ...The development of new technologically advanced products requires the contribution from a range of skills and disciplines, which are often difficult to fred within a single company or organization. Requirements establishment practices in Systems Engineering （SE）, while ensuring coordination of activities and tasks across the supply network, fall short when it comes to facilitate knowledge sharing and negotiation during early system design. Empirical observations show that when system-level requirements are not available or not mature enough, engineers dealing with the development of long lead-time sub-systems tend to target local optima, rather than opening up the design space. This phenomenon causes design teams to generate solutions that do not embody the best possible configuration for the overall system. The aim of this paper is to show how methodologies for value-driven design may address this issue, facilitating early stage design iterations and the resolution of early stage design trade-offs. The paper describes how such methodologies may help gathering and dispatching relevant knowledge about the ＇design intent＇ of a system to the cross-functional engineering teams, so to facilitate a more concurrent process for requirement elicitation in SE. The paper also describes EVOKE （Early Value Oriented design exploration with KnowledgE maturity）, a concept selection method that allows benchmarking design options at sub-system level on the base of value-related information communicated by the system integrators. The use of EVOKE is exemplified in an industrial case study related to the design of an aero-engine component. EVOKE＇s ability to raise awareness on the value contribution of early stage design concepts in the SE process has been further verified with industrial practitioners in ad-hoc design episodes.展开更多
The purpose of this article is to help small business persons who are in the market formicro-computers to select and use the specific product or service that will most effectively satisfytheir needs.This study is the ...The purpose of this article is to help small business persons who are in the market formicro-computers to select and use the specific product or service that will most effectively satisfytheir needs.This study is the development of a structure of representing system attributes in a formsuitable for a manageable decision model.This study uses“Descriptor”software package as a tooland uses the decision model of selecting a computer system and its vendor for an organization(buyer)to exemplify the application of“Descriptor”in decision processing.展开更多
A nonlinear model is proposed for effective adaptive control design. The model represents a natural way to describe input output nonlinear systems. A combined parameter off line estimation and structure detection al...A nonlinear model is proposed for effective adaptive control design. The model represents a natural way to describe input output nonlinear systems. A combined parameter off line estimation and structure detection algorithm is developed that can use an initial set of data. Then, an efficient model is obtained using orthogonal estimation with an error reduction test and other monitoring modifications. A recursive on line identification scheme is established based on the ELS algorithm to account for future time variations in the process of the parsimonious model. 展开更多
In machine learning, selecting useful features and rejecting redundant features is the prerequisite for better modeling and prediction. In this paper, we first study representative feature selection methods based on c...In machine learning, selecting useful features and rejecting redundant features is the prerequisite for better modeling and prediction. In this paper, we first study representative feature selection methods based on correlation analysis, and demonstrate that they do not work well for time series though they can work well for static systems. Then, theoretical analysis for linear time series is carried out to show why they fail. Based on these observations, we propose a new correlation-based feature selection method. Our main idea is that the features highly correlated with progressive response while lowly correlated with other features should be selected, and for groups of selected features with similar residuals, the one with a smaller number of features should be selected. For linear and nonlinear time series, the proposed method yields high accuracy in both feature selection and feature rejection.展开更多
基金the National Nature Science Foundation of China (60775047, 60402024)
文摘The support vector machine （SVM） is a novel machine learning method, which has the ability to approximate nonlinear functions with arbitrary accuracy. Setting parameters well is very crucial for SVM learning results and generalization ability, and now there is no systematic, general method for parameter selection. In this article, the SVM parameter selection for function approximation is regarded as a compound optimization problem and a mutative scale chaos optimization algorithm is employed to search for optimal paraxneter values. The chaos optimization algorithm is an effective way for global optimal and the mutative scale chaos algorithm could improve the search efficiency and accuracy. Several simulation examples show the sensitivity of the SVM parameters and demonstrate the superiority of this proposed method for nonlinear function approximation.
文摘Research on the role of sexual selection in the speciation process largely focuses on the diversifying role of mate choice. In particular, much attention has been drawn to the fact that population divergence in mate choice and in the male traits subject to choice directly can lead to assortative mating. However, male contest competition over mates also constitutes an important mechanism of sexual selection. We review recent empirical studies and argue that sexual selection through male contest competition can affect speciation in ways other than mate choice. For example, biases in aggression towards similar competitors can lead to disruptive and negative frequency-dependent selection on the traits used in contest competition in a similar way as competition for other types of limited resources. Moreover, male contest abilities often trade-off against other abilities such as parasite resistance, protection against predators and general stress tolerance. Populations experiencing different ecological condi- tions should therefore quickly diverge non-randomly in a number of traits including male contest abilities. In resource based breeding systems, a feedback loop between competitive ability and habitat use may lead to further population divergence. We discuss how population divergence in traits used in male contest competition can lead to the build up of reproductive isolation through a number of different pathways. Our main conclusion is that the role of male contest competition in speciation remains largely scientifically unexplored [Current Zoology 58 （3）： 493-509, 2012].
文摘The autonomous navigation of an Unmanned Aerial Vehicle(UAV)relies heavily on the navigation sensors.The UAV’s level of autonomy depends upon the various navigation systems,such as state measurement,mapping,and obstacle avoidance.Selecting the correct components is a critical part of the design process.However,this can be a particularly difficult task,especially for novices as there are several technologies and components available on the market,each with their own individual advantages and disadvantages.For example,satellite-based navigation components should be avoided when designing indoor UAVs.Incorporating them in the design brings no added value to the final product and will simply lead to increased cost and power consumption.Another issue is the number of vendors on the market,each trying to sell their hardware solutions which often incorporate similar technologies.The aim of this paper is to serve as a guide,proposing various methods to support the selection of fit-for-purpose technologies and components whilst avoiding system layout conflicts.The paper presents a study of the various navigation technologies and supports engineers in the selection of specific hardware solutions based on given requirements.The selection methods are based on easy-to-follow flow charts.A comparison of the various hardware components specifications is also included as part of this work.
文摘Currently,the Internet of Things(IoT)is revolutionizing communi-cation technology by facilitating the sharing of information between different physical devices connected to a network.To improve control,customization,flexibility,and reduce network maintenance costs,a new Software-Defined Network(SDN)technology must be used in this infrastructure.Despite the various advantages of combining SDN and IoT,this environment is more vulnerable to various attacks due to the centralization of control.Most methods to ensure IoT security are designed to detect Distributed Denial-of-Service(DDoS)attacks,but they often lack mechanisms to mitigate their severity.This paper proposes a Multi-Attack Intrusion Detection System(MAIDS)for Software-Defined IoT Networks(SDN-IoT).The proposed scheme uses two machine-learning algorithms to improve detection efficiency and provide a mechanism to prevent false alarms.First,a comparative analysis of the most commonly used machine-learning algorithms to secure the SDN was performed on two datasets:the Network Security Laboratory Knowledge Discovery in Databases(NSL-KDD)and the Canadian Institute for Cyberse-curity Intrusion Detection Systems(CICIDS2017),to select the most suitable algorithms for the proposed scheme and for securing SDN-IoT systems.The algorithms evaluated include Extreme Gradient Boosting(XGBoost),K-Nearest Neighbor(KNN),Random Forest(RF),Support Vector Machine(SVM),and Logistic Regression(LR).Second,an algorithm for selecting the best dataset for machine learning in Intrusion Detection Systems(IDS)was developed to enable effective comparison between the datasets used in the development of the security scheme.The results showed that XGBoost and RF are the best algorithms to ensure the security of SDN-IoT and to be applied in the proposed security system,with average accuracies of 99.88%and 99.89%,respectively.Furthermore,the proposed security scheme reduced the false alarm rate by 33.23%,which is a significant improvement over prevalent schemes.Finally,tests of the algorithm
文摘Traditional security systems are exposed to many various attacks,which represents a major challenge for the spread of the Internet in the future.Innovative techniques have been suggested for detecting attacks using machine learning and deep learning.The significant advantage of deep learning is that it is highly efficient,but it needs a large training time with a lot of data.Therefore,in this paper,we present a new feature reduction strategy based on Distributed Cumulative Histograms(DCH)to distinguish between dataset features to locate the most effective features.Cumulative histograms assess the dataset instance patterns of the applied features to identify the most effective attributes that can significantly impact the classification results.Three different models for detecting attacks using Convolutional Neural Network(CNN)and Long Short-Term Memory Network(LSTM)are also proposed.The accuracy test of attack detection using the hybrid model was 98.96%on the UNSW-NP15 dataset.The proposed model is compared with wrapper-based and filter-based Feature Selection(FS)models.The proposed model reduced classification time and increased detection accuracy.
基金National Key Research and Development Program of China(2020YFB1707700)National Natural Science Foundation of China(61972356,62036009)Fundamental Research Funds for the Provincial Universities of Zhejiang,China(RF-A2020001).
文摘Acoustic quality detection is vital in the manufactured products quality control field since it represents the conditions of machines or products.Recent work employed machine learning models in manufactured audio data to detect anomalous patterns.A major challenge is how to select applicable audio features to meliorate model’s accuracy and precision.To relax this challenge,we extract and analyze three audio feature types including Time Domain Feature,Frequency Domain Feature,and Cepstrum Feature to help identify the potential linear and non-linear relationships.In addition,we design a visual analysis system,namely AFExplorer,to assist data scientists in extracting audio features and selecting potential feature combinations.AFExplorer integrates four main views to present detailed distribution and relevance of the audio features,which helps users observe the impact of features visually in the feature selection.We perform the case study with AFExplore according to the ToyADMOS and MIMII Dataset to demonstrate the usability and effectiveness of the proposed system.
文摘The development of new technologically advanced products requires the contribution from a range of skills and disciplines, which are often difficult to fred within a single company or organization. Requirements establishment practices in Systems Engineering （SE）, while ensuring coordination of activities and tasks across the supply network, fall short when it comes to facilitate knowledge sharing and negotiation during early system design. Empirical observations show that when system-level requirements are not available or not mature enough, engineers dealing with the development of long lead-time sub-systems tend to target local optima, rather than opening up the design space. This phenomenon causes design teams to generate solutions that do not embody the best possible configuration for the overall system. The aim of this paper is to show how methodologies for value-driven design may address this issue, facilitating early stage design iterations and the resolution of early stage design trade-offs. The paper describes how such methodologies may help gathering and dispatching relevant knowledge about the ＇design intent＇ of a system to the cross-functional engineering teams, so to facilitate a more concurrent process for requirement elicitation in SE. The paper also describes EVOKE （Early Value Oriented design exploration with KnowledgE maturity）, a concept selection method that allows benchmarking design options at sub-system level on the base of value-related information communicated by the system integrators. The use of EVOKE is exemplified in an industrial case study related to the design of an aero-engine component. EVOKE＇s ability to raise awareness on the value contribution of early stage design concepts in the SE process has been further verified with industrial practitioners in ad-hoc design episodes.
文摘The purpose of this article is to help small business persons who are in the market formicro-computers to select and use the specific product or service that will most effectively satisfytheir needs.This study is the development of a structure of representing system attributes in a formsuitable for a manageable decision model.This study uses“Descriptor”software package as a tooland uses the decision model of selecting a computer system and its vendor for an organization(buyer)to exemplify the application of“Descriptor”in decision processing.
文摘A nonlinear model is proposed for effective adaptive control design. The model represents a natural way to describe input output nonlinear systems. A combined parameter off line estimation and structure detection algorithm is developed that can use an initial set of data. Then, an efficient model is obtained using orthogonal estimation with an error reduction test and other monitoring modifications. A recursive on line identification scheme is established based on the ELS algorithm to account for future time variations in the process of the parsimonious model.
文摘In machine learning, selecting useful features and rejecting redundant features is the prerequisite for better modeling and prediction. In this paper, we first study representative feature selection methods based on correlation analysis, and demonstrate that they do not work well for time series though they can work well for static systems. Then, theoretical analysis for linear time series is carried out to show why they fail. Based on these observations, we propose a new correlation-based feature selection method. Our main idea is that the features highly correlated with progressive response while lowly correlated with other features should be selected, and for groups of selected features with similar residuals, the one with a smaller number of features should be selected. For linear and nonlinear time series, the proposed method yields high accuracy in both feature selection and feature rejection.
