This paper firstly elaborated current situation of agricultural resources and ecological environment protection of China,and pointed out that circular agriculture is an inevitable choice for sustainable development of...This paper firstly elaborated current situation of agricultural resources and ecological environment protection of China,and pointed out that circular agriculture is an inevitable choice for sustainable development of agriculture and effective approach for low-carbon economic development. Then,it analyzed mode and benefit of typical circular agriculture,such as ecological orchard in southeast hills,sightseeing ecological agricultural garden,and crop straw use. Finally,it came up with countermeasures for further developing circular agriculture,including integrating building system and establishing suitable mode,formulating production standard and ensuring food safety,strengthening technological innovation,and supporting sustainable development.展开更多
The Red Soil Hilly Region in South China, where there is a high capacity of carbon(C), and the land use and vegetation cover change greatly, is an important ecological area in the world, and has an important impact on...The Red Soil Hilly Region in South China, where there is a high capacity of carbon(C), and the land use and vegetation cover change greatly, is an important ecological area in the world, and has an important impact on the global carbon cycle and the seasonal fluctuation of atmospheric CO_2. To better evaluate the effects of reclamation systems in orchards converted from grasslands on soil carbon sequestration, we investigated soil organic carbon(SOC) content and stable C isotope(δ^(13)C)composition in three nectarine orchards at Yuchi Experimental Station in South China. Compared with the sloping clean tillage orchard and terraced clean tillage orchard, SOC content in the terraced orchard with grass cover was increased by 14.90% to 38.49%, and 7.40% to 15.33%, respectively. During the 14 years after orchard establishment, the soil organic matter sources influenced both δ^(13)C distribution with depth and carbon replacement. SOC turnover of the upper soil layer in the terraced orchard with grass cover(a mean 63.05% of replacement in the 20 cm after 14 years) was 1.59 and 1.41 times larger than that of the sloping clean tillage orchard and terraced clean tillage orchard under subtropical conditions, respectively. The equilibrium value of soil organic carbon in the three treatments ranged from 16.067 to 25.608 g/kg under the experimental conditions. The equilibrium value of soil organic carbon in the surface layer under grass cover was 54.801 t/hm^2, and the carbon sequestration potential was 24.695 1 t/hm^2.展开更多
Soil erosion occurred in orchards has often attracted extensive attentions from the society with environmental considerations,as orchard is one of major methods of agricultural production in China.In the hilly red soi...Soil erosion occurred in orchards has often attracted extensive attentions from the society with environmental considerations,as orchard is one of major methods of agricultural production in China.In the hilly red soil region of China,many orchards are established on slope lands with a lack of grass covers,leading to severe soil losses.In order to mitigate this common environmental problem and evaluate the efficiency of erosion-control approaches,four treatments were set in field plots in this study,including terraced peach orchard with Arachis pintoi cv.Amarillo as mulch and Paspalum natatu and Digitariasmutsii as hedgerows(TTM),terraced peach orchard without conservation measures(TTW),sloping peach orchard with A.pintoi as mulch and P.natatu and Digitariasmutsii as hedgerows(STM),and sloping peach orchard without conservation measures(STW).The surface runoff,sediment yields and the contents of soil nutrients and organic carbon were monitored in the four treatments and the comprehensive eco-service benefits were further evaluated.The results indicate that available phosphorus(AP),available potassium(AK),total nitrogen(TN) and organic matter(OM) in the soils of the TTM treatments and STM were significantly higher than those of the treatments TTW and STW,suggesting positive effects of the vegetation covers on the soil nutrients.Mean annual surface runoff and coefficient ranged from 0.86 to 34.79 m^3,and 0.007 to 0.282,respectively;the treatment TTM exhibited the best water conservation benefits and the treatment STW was the worst.Soil erosion modulus of the plots were 0-28.76t/hm^2 per year in average,and the treatments TTM and STM reduced significantly soil loss in comparison of the treatments TTW and STW;(d) total organic carbon in the vegetation covers ranged from 130.23 to 195.93 kg per year,and that for TTM and STM treatment significantly higher than TTW and STW treatment;comprehensive eco-service values of the orchards were evaluated considering all the factors including water conservation,soil fertility conservation,CO_2 fixation and O_2 supply,ranging from 563.35 $/y to 765.51 $/y.As expected,the treatments TTM and STM had significantly greater eco-service values than the treatments TTW and STW.In summary,we concluded that terraced orchard with A.pintoi as live mulch plus Paspalum natatu and Digitariasmutsii as hedgerows is a highly sustainable land use practice for the slope lands in red soil hilly region of China.展开更多
Water management in proton exchange membrane fuel cells(PEMFC)is a topic of great importance for the optimization of these systems.Effective proton conductivity calls for moderate moisture content in the membrane,whil...Water management in proton exchange membrane fuel cells(PEMFC)is a topic of great importance for the optimization of these systems.Effective proton conductivity calls for moderate moisture content in the membrane,while uneven water distribution can lead to instability of the whole flow field,thereby decreasing the performance of the fuel cell.In the present study,a simplified two-tier hybrid structure is used to investigate the impact of the dynamic behavior of liquid water on the current density of the PEMFC.Simulation results show that water droplets attached to wall sides tend to increase current density.Visualization experiments confirm the existence of liquid droplets and the enhancement of current density,while indicating that the best performance and stability of fuel cell are attained for a cathode air flow rate of 300 ml/min.展开更多
Carbon neutralization has been introduced as a long-term policy to control global warming and climate change.As plant photosynthesis produces the most abundant lignocellulosic biomass on Earth,its conversion to biofue...Carbon neutralization has been introduced as a long-term policy to control global warming and climate change.As plant photosynthesis produces the most abundant lignocellulosic biomass on Earth,its conversion to biofuels and bioproducts is considered a promising solution for reducing the net carbon release.However,natural lignocellulose recalcitrance crucially results in a costly biomass process along with secondary waste liberation.By updating recent advances in plant biotechnology,biomass engineering,and carbon nanotechnology,this study proposes a novel strategy that integrates the genetic engineering of bioenergy crops with green-like biomass processing for cost-effective biofuel conversion and high-value bioproduction.By selecting key genes and appropriate genetic manipulation approaches for precise lignocellulose modification,this study highlights the desirable genetic site mutants and transgenic lines that are raised in amorphous regions and inner broken chains account for high-density/length-reduced cellulose nanofiber assembly in situ.Since the amorphous regions and inner-broken chains of lignocellulose substrates are defined as the initial breakpoints for enhancing biochemical,chemical,and thermochemical conversions,desirable cellulose nanofibers can be employed to achieve nearcomplete biomass enzymatic saccharification for maximizing biofuels or high-quality biomaterials,even under cost-effective and green-like biomass processes in vitro.This study emphasizes the optimal thermal conversion for generating high-performance nanocarbons by combining appropriate nanomaterials generated from diverse lignocellulose resources.Therefore,this study provides a perspective on the potential of green carbon productivity as a part of the fourth industrial revolution.展开更多
Topological edge states have an important role in optical modulation with potential applications in wavelength division multiplexers(WDMs).In this paper,2D photonic crystals(PCs)with different rotation angles are comb...Topological edge states have an important role in optical modulation with potential applications in wavelength division multiplexers(WDMs).In this paper,2D photonic crystals(PCs)with different rotation angles are combined to generate topological edge states.We reveal the relationship between the edge states and the rotation parameters of PCs,and further propose a WDM to realize the application of adjustable beams.Our findings successfully reveal the channel selectivity for optical transmission and provide a flexible way to promote the development of topological photonic devices.展开更多
In order to detect the multiple avian influenza viruses(AIVs)rapidly,specifically and sensitively,a LabVIEW and microelectrode array-based impedance biosensor was developed and demonstrated.A laptop with LabVIEW softw...In order to detect the multiple avian influenza viruses(AIVs)rapidly,specifically and sensitively,a LabVIEW and microelectrode array-based impedance biosensor was developed and demonstrated.A laptop with LabVIEW software was used to generate excitation signals at different frequencies with an audio card and measure the impedance of target viruses through a data acquisition card.The audio card of the laptop was used as a function generator,while a data acquisition card was used for data communication.A virtual instrument was programmed with LabVIEW to provide a platform for impedance measurement,data processing,and control.Six interdigitated microelectrodes were placed at the bottom of six wells on a microplate to form six sensors for different AIVs and controls.Then,AIV specific ligands were immobilized on the microelectrode surface to capture target viruses.To enhance the sensitivity,AIV specific aptamers conjugated gold nanoparticles and thiocyanuric acid were employed to form a network structure and used as an amplifier.Results of the measured impedance were compared with a commercial IM6 impedance analyzer,and the error was less than 5%.The developed biosensor was portable with the sensitivity and specificity for applications to on-site or in-field rapid screening of avian influenza viruses.展开更多
Adversarial Malware Example(AME)-based adversarial training can effectively enhance the robustness of Machine Learning(ML)-based malware detectors against AME.AME quality is a key factor to the robustness enhancement....Adversarial Malware Example(AME)-based adversarial training can effectively enhance the robustness of Machine Learning(ML)-based malware detectors against AME.AME quality is a key factor to the robustness enhancement.Generative Adversarial Network(GAN)is a kind of AME generation method,but the existing GAN-based AME generation methods have the issues of inadequate optimization,mode collapse and training instability.In this paper,we propose a novel approach(denote as LSGAN-AT)to enhance ML-based malware detector robustness against Adversarial Examples,which includes LSGAN module and AT module.LSGAN module can generate more effective and smoother AME by utilizing brand-new network structures and Least Square(LS)loss to optimize boundary samples.AT module makes adversarial training using AME generated by LSGAN to generate ML-based Robust Malware Detector(RMD).Extensive experiment results validate the better transferability of AME in terms of attacking 6 ML detectors and the RMD transferability in terms of resisting the MalGAN black-box attack.The results also verify the performance of the generated RMD in the recognition rate of AME.展开更多
Adversarial Malware Example(AME)-based adversarial training can effectively enhance the robustness of Machine Learning(ML)-based malware detectors against AME.AME quality is a key factor to the robustness enhancement....Adversarial Malware Example(AME)-based adversarial training can effectively enhance the robustness of Machine Learning(ML)-based malware detectors against AME.AME quality is a key factor to the robustness enhancement.Generative Adversarial Network(GAN)is a kind of AME generation method,but the existing GAN-based AME generation methods have the issues of inadequate optimization,mode collapse and training instability.In this paper,we propose a novel approach(denote as LSGAN-AT)to enhance ML-based malware detector robustness against Adversarial Examples,which includes LSGAN module and AT module.LSGAN module can generate more effective and smoother AME by utilizing brand-new network structures and Least Square(LS)loss to optimize boundary samples.AT module makes adversarial training using AME generated by LSGAN to generate ML-based Robust Malware Detector(RMD).Extensive experiment results validate the better transferability of AME in terms of attacking 6 ML detectors and the RMD transferability in terms of resisting the MalGAN black-box attack.The results also verify the performance of the generated RMD in the recognition rate of AME.展开更多
Deep learning(DL)has exhibited its exceptional performance in fields like intrusion detection.Various augmentation methods have been proposed to improve data quality and eventually to enhance the performance of DL mod...Deep learning(DL)has exhibited its exceptional performance in fields like intrusion detection.Various augmentation methods have been proposed to improve data quality and eventually to enhance the performance of DL models.However,the classic augmentation methods cannot be applied to those DL models which exploit the system-call sequences to detect intrusion.Previously,the seq2seq model has been explored to augment system-call sequences.Following this work,we propose a gated convolutional neural network(GCNN)model to thoroughly extract the potential information of augmented sequences.Also,in order to enhance themodel’s robustness,we adopt adversarial training to reduce the impact of adversarial examples on the model.Adversarial examples used in adversarial training are generated by the proposed adversarial sequence generation algorithm.The experimental results on different verified models show that GCNN model can better obtain the potential information of the augmented data and achieve the best performance.Furthermore,GCNN with adversarial training can enhance robustness significantly.展开更多
Deep learning(DL)has exhibited its exceptional performance in fields like intrusion detection.Various augmentation methods have been proposed to improve data quality and eventually to enhance the performance of DL mod...Deep learning(DL)has exhibited its exceptional performance in fields like intrusion detection.Various augmentation methods have been proposed to improve data quality and eventually to enhance the performance of DL models.However,the classic augmentation methods cannot be applied to those DL models which exploit the system-call sequences to detect intrusion.Previously,the seq2seq model has been explored to augment system-call sequences.Following this work,we propose a gated convolutional neural network(GCNN)model to thoroughly extract the potential information of augmented sequences.Also,in order to enhance themodel’s robustness,we adopt adversarial training to reduce the impact of adversarial examples on the model.Adversarial examples used in adversarial training are generated by the proposed adversarial sequence generation algorithm.The experimental results on different verified models show that GCNN model can better obtain the potential information of the augmented data and achieve the best performance.Furthermore,GCNN with adversarial training can enhance robustness significantly.展开更多
Tricho-dento-osseous(TDO)syndrome is a rare autosomal dominant disease resulting from distal-less homeobox 3(DLX3)mutation.1,2 Accumulative bone density in alveolar bone is a clinically favorable phenotype for TDO pat...Tricho-dento-osseous(TDO)syndrome is a rare autosomal dominant disease resulting from distal-less homeobox 3(DLX3)mutation.1,2 Accumulative bone density in alveolar bone is a clinically favorable phenotype for TDO patients.However,the limited number of bone marrow mesenchymal stem cells(BMSCs)in TDO patients restricts their application.展开更多
基金Supported by the National Science and Technology Project in the Twelfth Five-year Plan Period(2012BAD14B15)
文摘This paper firstly elaborated current situation of agricultural resources and ecological environment protection of China,and pointed out that circular agriculture is an inevitable choice for sustainable development of agriculture and effective approach for low-carbon economic development. Then,it analyzed mode and benefit of typical circular agriculture,such as ecological orchard in southeast hills,sightseeing ecological agricultural garden,and crop straw use. Finally,it came up with countermeasures for further developing circular agriculture,including integrating building system and establishing suitable mode,formulating production standard and ensuring food safety,strengthening technological innovation,and supporting sustainable development.
基金Supported by Science and Technology Program of Fujian Province(2017R1016-4)Natural Science Foundation of Fujian Province(2017J01072)
文摘The Red Soil Hilly Region in South China, where there is a high capacity of carbon(C), and the land use and vegetation cover change greatly, is an important ecological area in the world, and has an important impact on the global carbon cycle and the seasonal fluctuation of atmospheric CO_2. To better evaluate the effects of reclamation systems in orchards converted from grasslands on soil carbon sequestration, we investigated soil organic carbon(SOC) content and stable C isotope(δ^(13)C)composition in three nectarine orchards at Yuchi Experimental Station in South China. Compared with the sloping clean tillage orchard and terraced clean tillage orchard, SOC content in the terraced orchard with grass cover was increased by 14.90% to 38.49%, and 7.40% to 15.33%, respectively. During the 14 years after orchard establishment, the soil organic matter sources influenced both δ^(13)C distribution with depth and carbon replacement. SOC turnover of the upper soil layer in the terraced orchard with grass cover(a mean 63.05% of replacement in the 20 cm after 14 years) was 1.59 and 1.41 times larger than that of the sloping clean tillage orchard and terraced clean tillage orchard under subtropical conditions, respectively. The equilibrium value of soil organic carbon in the three treatments ranged from 16.067 to 25.608 g/kg under the experimental conditions. The equilibrium value of soil organic carbon in the surface layer under grass cover was 54.801 t/hm^2, and the carbon sequestration potential was 24.695 1 t/hm^2.
文摘Soil erosion occurred in orchards has often attracted extensive attentions from the society with environmental considerations,as orchard is one of major methods of agricultural production in China.In the hilly red soil region of China,many orchards are established on slope lands with a lack of grass covers,leading to severe soil losses.In order to mitigate this common environmental problem and evaluate the efficiency of erosion-control approaches,four treatments were set in field plots in this study,including terraced peach orchard with Arachis pintoi cv.Amarillo as mulch and Paspalum natatu and Digitariasmutsii as hedgerows(TTM),terraced peach orchard without conservation measures(TTW),sloping peach orchard with A.pintoi as mulch and P.natatu and Digitariasmutsii as hedgerows(STM),and sloping peach orchard without conservation measures(STW).The surface runoff,sediment yields and the contents of soil nutrients and organic carbon were monitored in the four treatments and the comprehensive eco-service benefits were further evaluated.The results indicate that available phosphorus(AP),available potassium(AK),total nitrogen(TN) and organic matter(OM) in the soils of the TTM treatments and STM were significantly higher than those of the treatments TTW and STW,suggesting positive effects of the vegetation covers on the soil nutrients.Mean annual surface runoff and coefficient ranged from 0.86 to 34.79 m^3,and 0.007 to 0.282,respectively;the treatment TTM exhibited the best water conservation benefits and the treatment STW was the worst.Soil erosion modulus of the plots were 0-28.76t/hm^2 per year in average,and the treatments TTM and STM reduced significantly soil loss in comparison of the treatments TTW and STW;(d) total organic carbon in the vegetation covers ranged from 130.23 to 195.93 kg per year,and that for TTM and STM treatment significantly higher than TTW and STW treatment;comprehensive eco-service values of the orchards were evaluated considering all the factors including water conservation,soil fertility conservation,CO_2 fixation and O_2 supply,ranging from 563.35 $/y to 765.51 $/y.As expected,the treatments TTM and STM had significantly greater eco-service values than the treatments TTW and STW.In summary,we concluded that terraced orchard with A.pintoi as live mulch plus Paspalum natatu and Digitariasmutsii as hedgerows is a highly sustainable land use practice for the slope lands in red soil hilly region of China.
基金by the National Natural Science Foundation of China(51175472)the Natural Science Foundation of Zhejiang Province(LQ20E060008)the Foundation of Department of Education of Zhejiang Province(Y201737452).
文摘Water management in proton exchange membrane fuel cells(PEMFC)is a topic of great importance for the optimization of these systems.Effective proton conductivity calls for moderate moisture content in the membrane,while uneven water distribution can lead to instability of the whole flow field,thereby decreasing the performance of the fuel cell.In the present study,a simplified two-tier hybrid structure is used to investigate the impact of the dynamic behavior of liquid water on the current density of the PEMFC.Simulation results show that water droplets attached to wall sides tend to increase current density.Visualization experiments confirm the existence of liquid droplets and the enhancement of current density,while indicating that the best performance and stability of fuel cell are attained for a cathode air flow rate of 300 ml/min.
基金supported by the National Natural Science Foundation of China(32170268 to L.P)the National 111 Project of the Ministry of Education of China(BP0820035 to L.P,D17009 to J.T)+1 种基金the Initiative Grant of Hubei University of Technology for High-level Talents(GCC20230001 to L.P)the Shandong Energy Institute,China(SEI I202142 to C.F).
文摘Carbon neutralization has been introduced as a long-term policy to control global warming and climate change.As plant photosynthesis produces the most abundant lignocellulosic biomass on Earth,its conversion to biofuels and bioproducts is considered a promising solution for reducing the net carbon release.However,natural lignocellulose recalcitrance crucially results in a costly biomass process along with secondary waste liberation.By updating recent advances in plant biotechnology,biomass engineering,and carbon nanotechnology,this study proposes a novel strategy that integrates the genetic engineering of bioenergy crops with green-like biomass processing for cost-effective biofuel conversion and high-value bioproduction.By selecting key genes and appropriate genetic manipulation approaches for precise lignocellulose modification,this study highlights the desirable genetic site mutants and transgenic lines that are raised in amorphous regions and inner broken chains account for high-density/length-reduced cellulose nanofiber assembly in situ.Since the amorphous regions and inner-broken chains of lignocellulose substrates are defined as the initial breakpoints for enhancing biochemical,chemical,and thermochemical conversions,desirable cellulose nanofibers can be employed to achieve nearcomplete biomass enzymatic saccharification for maximizing biofuels or high-quality biomaterials,even under cost-effective and green-like biomass processes in vitro.This study emphasizes the optimal thermal conversion for generating high-performance nanocarbons by combining appropriate nanomaterials generated from diverse lignocellulose resources.Therefore,this study provides a perspective on the potential of green carbon productivity as a part of the fourth industrial revolution.
基金National Key Research and Development Program of China(2022YFE0122300)National Natural Science Foundation of China(11811530052,12004425,1211101294,62105126)+2 种基金State Key Laboratory of Millimeter Waves(K202105,K202238)Intergovernmental Science and Technology Regular Meeting Exchange Project of Ministry of Science and Technology of China(CB02-20)Natural Science Foundation of Jiangsu Province(BK20200630)。
文摘Topological edge states have an important role in optical modulation with potential applications in wavelength division multiplexers(WDMs).In this paper,2D photonic crystals(PCs)with different rotation angles are combined to generate topological edge states.We reveal the relationship between the edge states and the rotation parameters of PCs,and further propose a WDM to realize the application of adjustable beams.Our findings successfully reveal the channel selectivity for optical transmission and provide a flexible way to promote the development of topological photonic devices.
文摘In order to detect the multiple avian influenza viruses(AIVs)rapidly,specifically and sensitively,a LabVIEW and microelectrode array-based impedance biosensor was developed and demonstrated.A laptop with LabVIEW software was used to generate excitation signals at different frequencies with an audio card and measure the impedance of target viruses through a data acquisition card.The audio card of the laptop was used as a function generator,while a data acquisition card was used for data communication.A virtual instrument was programmed with LabVIEW to provide a platform for impedance measurement,data processing,and control.Six interdigitated microelectrodes were placed at the bottom of six wells on a microplate to form six sensors for different AIVs and controls.Then,AIV specific ligands were immobilized on the microelectrode surface to capture target viruses.To enhance the sensitivity,AIV specific aptamers conjugated gold nanoparticles and thiocyanuric acid were employed to form a network structure and used as an amplifier.Results of the measured impedance were compared with a commercial IM6 impedance analyzer,and the error was less than 5%.The developed biosensor was portable with the sensitivity and specificity for applications to on-site or in-field rapid screening of avian influenza viruses.
基金The research of J.Wang,X.Chang,Y.Wang and J.Zhang was supported in part by Project supported by Chinese National Key Laboratory of Science and Technology on Information System Security and National Natural Science Foundation of China under Grant No.U1836105The research of R.J.Rodriguez and X.Chang has been supported in part by the University of Zaragoza and the Fundacion Ibercaja under Grant JIUZ-2020-TIC-08The research of R.J.Rodriguez has also been supported in part by the University,Industry and Innovation Department of the Aragonese Government under Programa de Proyectos Estrategicos de Grupos de Investigacidn(DisCo research group,ref.T21-20R).
文摘Adversarial Malware Example(AME)-based adversarial training can effectively enhance the robustness of Machine Learning(ML)-based malware detectors against AME.AME quality is a key factor to the robustness enhancement.Generative Adversarial Network(GAN)is a kind of AME generation method,but the existing GAN-based AME generation methods have the issues of inadequate optimization,mode collapse and training instability.In this paper,we propose a novel approach(denote as LSGAN-AT)to enhance ML-based malware detector robustness against Adversarial Examples,which includes LSGAN module and AT module.LSGAN module can generate more effective and smoother AME by utilizing brand-new network structures and Least Square(LS)loss to optimize boundary samples.AT module makes adversarial training using AME generated by LSGAN to generate ML-based Robust Malware Detector(RMD).Extensive experiment results validate the better transferability of AME in terms of attacking 6 ML detectors and the RMD transferability in terms of resisting the MalGAN black-box attack.The results also verify the performance of the generated RMD in the recognition rate of AME.
基金Chinese National Key Laboratory of Science and Technology on Information System Security and National Natural Science Foundation of China under Grant No.U1836105The research of R.J.Rodríguez and X.Chang has been supported in part by the University of Zaragoza and the Fundación Ibercaja under Grant JIUZ-2020-TIC-08The research of R.J.Rodríguez has also been supported in part by the University,Industry and Innovation Department of the Aragonese Government under Programa de Proyectos Estratégicos de Grupos de Investigación(DisCo research group,ref.T21-20R).
文摘Adversarial Malware Example(AME)-based adversarial training can effectively enhance the robustness of Machine Learning(ML)-based malware detectors against AME.AME quality is a key factor to the robustness enhancement.Generative Adversarial Network(GAN)is a kind of AME generation method,but the existing GAN-based AME generation methods have the issues of inadequate optimization,mode collapse and training instability.In this paper,we propose a novel approach(denote as LSGAN-AT)to enhance ML-based malware detector robustness against Adversarial Examples,which includes LSGAN module and AT module.LSGAN module can generate more effective and smoother AME by utilizing brand-new network structures and Least Square(LS)loss to optimize boundary samples.AT module makes adversarial training using AME generated by LSGAN to generate ML-based Robust Malware Detector(RMD).Extensive experiment results validate the better transferability of AME in terms of attacking 6 ML detectors and the RMD transferability in terms of resisting the MalGAN black-box attack.The results also verify the performance of the generated RMD in the recognition rate of AME.
基金This work was supported in part by the Fundamental Research Funds for the Central Universities of China under Grants 2019YJS049。
文摘Deep learning(DL)has exhibited its exceptional performance in fields like intrusion detection.Various augmentation methods have been proposed to improve data quality and eventually to enhance the performance of DL models.However,the classic augmentation methods cannot be applied to those DL models which exploit the system-call sequences to detect intrusion.Previously,the seq2seq model has been explored to augment system-call sequences.Following this work,we propose a gated convolutional neural network(GCNN)model to thoroughly extract the potential information of augmented sequences.Also,in order to enhance themodel’s robustness,we adopt adversarial training to reduce the impact of adversarial examples on the model.Adversarial examples used in adversarial training are generated by the proposed adversarial sequence generation algorithm.The experimental results on different verified models show that GCNN model can better obtain the potential information of the augmented data and achieve the best performance.Furthermore,GCNN with adversarial training can enhance robustness significantly.
基金supported in part by the Fundamental Research Funds for the Central Universities of China under Grants 2019YJS049。
文摘Deep learning(DL)has exhibited its exceptional performance in fields like intrusion detection.Various augmentation methods have been proposed to improve data quality and eventually to enhance the performance of DL models.However,the classic augmentation methods cannot be applied to those DL models which exploit the system-call sequences to detect intrusion.Previously,the seq2seq model has been explored to augment system-call sequences.Following this work,we propose a gated convolutional neural network(GCNN)model to thoroughly extract the potential information of augmented sequences.Also,in order to enhance themodel’s robustness,we adopt adversarial training to reduce the impact of adversarial examples on the model.Adversarial examples used in adversarial training are generated by the proposed adversarial sequence generation algorithm.The experimental results on different verified models show that GCNN model can better obtain the potential information of the augmented data and achieve the best performance.Furthermore,GCNN with adversarial training can enhance robustness significantly.
基金supported by the National Nature Science Foundation of China(No.81970920,81900983)the Natural Science Foundation of Beijing Municipality,China(No.7232218)the Shanghai Science and Technology Committee Youth Sailing Program(China)(No.19YF1442500).
文摘Tricho-dento-osseous(TDO)syndrome is a rare autosomal dominant disease resulting from distal-less homeobox 3(DLX3)mutation.1,2 Accumulative bone density in alveolar bone is a clinically favorable phenotype for TDO patients.However,the limited number of bone marrow mesenchymal stem cells(BMSCs)in TDO patients restricts their application.