Water conservation initiatives satisfy the demand for water supply,electricity generation,irrigation,and flood control.While helping humanity,they have also altered the ecosystem of natural rivers,impacted river ecolo...Water conservation initiatives satisfy the demand for water supply,electricity generation,irrigation,and flood control.While helping humanity,they have also altered the ecosystem of natural rivers,impacted river ecology,disrupted river continuity,and jeopardized the existence of aquatic creatures in rivers.Studying the impacts of dam construction on rivers can enhance our knowledge of river ecological and environmental concerns and help sustain the health of river ecosystems,thereby realizing the harmony between humans and water in both theoretical and practical aspects.This study used bibliometrics and constructed an author-keyword 2-mode matrix network using Co-Occurrence software to identify the hotspots and research trend in eco-hydrology of dammed rivers.We identified‘FLOW’‘SEDIMENT’‘QUALITY’and‘MODEL’as the research hotspots in the ecological impact of dammed rivers,and combined the related literatures,we highlight the research progress in the four directions.Then the research shortcomings and prospect were discussed,including strengthening the monitoring and analysis of critical ecological variables,enhancing the hydrological monitoring density for small rivers,strengthening the research of relationship between eutrophication and zooplankton,establishing multiscale approaches,and combining multi-sources information technologies to improve parameter accuracy in the model research.展开更多
Organic semiconductor materials have shown unique advantages in the development of optoelectronic devices due to their ease of preparation,low cost,lightweight,and flexibility.In this work,we explored the application ...Organic semiconductor materials have shown unique advantages in the development of optoelectronic devices due to their ease of preparation,low cost,lightweight,and flexibility.In this work,we explored the application of the organic semiconductor Y6-1O single crystal in photodetection devices.Firstly,Y6-1O single crystal material was prepared on a silicon substrate using solution droplet casting method.The optical properties of Y6-1O material were characterized by polarized optical microscopy,fluorescence spectroscopy,etc.,confirming its highly single crystalline performance and emission properties in the near-infrared region.Phototransistors based on Y6-1O materials with different thicknesses were then fabricated and tested.It was found that the devices exhibited good visible to near-infrared photoresponse,with the maximum photoresponse in the near-infrared region at 785 nm.The photocurrent on/off ratio reaches 10^(2),and photoresponsivity reaches 16 mA/W.It was also found that the spectral response of the device could be regulated by gate voltage as well as the material thickness,providing important conditions for optimizing the performance of near-infrared photodetectors.This study not only demonstrates the excellent performance of organic phototransistors based on Y6-1O single crystal material in near-infrared detection but also provides new ideas and directions for the future development of infrared detectors.展开更多
The Quadric Error Metrics(QEM)algorithm is a widely used method for mesh simplification;however,it often struggles to preserve high-frequency geometric details,leading to the loss of salient features.To address this l...The Quadric Error Metrics(QEM)algorithm is a widely used method for mesh simplification;however,it often struggles to preserve high-frequency geometric details,leading to the loss of salient features.To address this limitation,we propose the Salient Feature Sampling Points-based QEM(SFSP-QEM)—also referred to as the Deep Learning-Based Salient Feature-Preserving Algorithm for Mesh Simplification—which incorporates a Salient Feature-Preserving Point Sampler(SFSP).This module leverages deep learning techniques to prioritize the preservation of key geometric features during simplification.Experimental results demonstrate that SFSP-QEM significantly outperforms traditional QEM in preserving geometric details.Specifically,for general models from the Stanford 3D Scanning Repository,which represent typical mesh structures used in mesh simplification benchmarks,the Hausdorff distance of simplified models using SFSP-QEM is reduced by an average of 46.58% compared to those simplified using traditional QEM.In customized models such as the Zigong Lantern used in cultural heritage preservation,SFSP-QEM achieves an average reduction of 28.99% in Hausdorff distance.Moreover,the running time of this method is only 6%longer than that of traditional QEM while significantly improving the preservation of geometric details.These results demonstrate that SFSP-QEMis particularly effective for applications requiring high-fidelity simplification while retaining critical features.展开更多
Addressing the current challenges in transforming pixel displacement into physical displacement in visual monitoring technologies,as well as the inability to achieve precise full-field monitoring,this paper proposes a...Addressing the current challenges in transforming pixel displacement into physical displacement in visual monitoring technologies,as well as the inability to achieve precise full-field monitoring,this paper proposes a method for identifying the structural dynamic characteristics of wind turbines based on visual monitoring data fusion.Firstly,the Lucas-Kanade Tomasi(LKT)optical flow method and a multi-region of interest(ROI)monitoring structure are employed to track pixel displacements,which are subsequently subjected to band pass filtering and resampling operations.Secondly,the actual displacement time history is derived through double integration of the acquired acceleration data and subsequent band pass filtering.The scale factor is obtained by applying the least squares method to compare the visual displacement with the displacement derived from double integration of the acceleration data.Based on this,the multi-point displacement time histories under physical coordinates are obtained using the vision data and the scale factor.Subsequently,when visual monitoring of displacements becomes impossible due to issues such as image blurring or lens occlusion,the structural vibration equation and boundary condition constraints,among other key parameters,are employed to predict the displacements at unknown monitoring points,thereby enabling full-field displacement monitoring and dynamic characteristic testing of the structure.Finally,a small-scale shaking table test was conducted on a simulated wind turbine structure undergoing shutdown to validate the dynamic characteristics of the proposed method through test verification.The research results indicate that the proposed method achieves a time-domain error within the submillimeter range and a frequency-domain accuracy of over 99%,effectively monitoring the full-field structural dynamic characteristics of wind turbines and providing a basis for the condition assessment of wind turbine structures.展开更多
Dunhuang medicine is an important part of Dunhuang studies,boasting rich and comprehensive connotations.It records more than 1,000 Dunhuang medical prescriptions,involving internal,external,gynecological,pediatric,Bud...Dunhuang medicine is an important part of Dunhuang studies,boasting rich and comprehensive connotations.It records more than 1,000 Dunhuang medical prescriptions,involving internal,external,gynecological,pediatric,Buddhist,and Daoist medicine,and has demonstrated good clinical effects.However,the mechanism of action of relevant Dunhuang medical prescriptions is still unclear,existing research lacks systematic review and summarization,which has limited their further development.At the same time,the inheritance,innovation,and transformation of Dunhuang medicine are critical issues for the development of Dunhuang medicine,which has important guiding significance for the future development of Dunhuang medicine.Therefore,this study systematically summarizes the experimental research progress of Dunhuang medical prescriptions[except for those contained in Fu Xing Jue Zang Fu Yong Yao Fa Yao(《辅行诀脏腑用药法要》The Guideline to Use Medicines for Zang-fu)],and seven such prescriptions were selected based on three criteria:well-preserved texts,no prior transmission to the outside world,and having extensive research and clinical application over the past decade.The findings indicate that this type of prescription is applicable to a broad spectrum of diseases and has a promising application prospect in health preservation and disease prevention,as it exerts therapeutic effects through multiple targets and pathways.Based on this,specific strategies for the transformation of Dunhuang characteristic prescriptions were proposed from three aspects:inheritance,innovative development,and transformation strategies,aiming to provide insights for the future development of Dunhuang medical prescriptions.展开更多
Background Cotton is an important crop providing the most natural fibers all over the world. The cotton genomics community has utilized whole genome sequencing data to construct an elite gene pool in which functional ...Background Cotton is an important crop providing the most natural fibers all over the world. The cotton genomics community has utilized whole genome sequencing data to construct an elite gene pool in which functional genes are related to agronomic traits. However, the functional validation of these genes is hindered by time-consuming and inefficient genetic transformation methods. Thus, establishing a transient transformation system of high efficiency is necessary for cotton genomics.Results To improve the efficiency of transient transformation, we used the protoplasts isolated from the etiolated cotyledon as recipient. The enzymatic digestion buffer comprised 1.5%(w/v) cellulase, 0.75%(w/v) macerozyme, and 1% hemicellulase, osmotically buffered with 0.4 mol·L^(-1) mannitol. After 5 h of dark incubation at 25℃, uniform cotton protoplasts were successfully isolated with a yield of 4.6 × 10^(6) protoplasts per gram(fresh weight) and 95% viability. We incubated 100 μL protoplasts(2.5 × 10^(5)·m L^(-1)) with 15 μg plasmid in the solution of 0.4 mol·L^(-1) mannitol and 40% PEG 4000 for 15 min, ultimately achieving an optimal transient transfection efficiency of 71.47%.Conclusions This transient system demonstrated effective utility in cellular biology research through successful applications in subcellular localization analyses, bimolecular fluorescence complementation(Bi FC) verification, and prime editing vector validation. Through systematic optimization, we established an efficient and expedited protoplast-based transient transformation system and successfully applied this platform to cotton functional genomics studies.展开更多
To reveal the effects of environmental and loading conditions, as well as asphalt properties on the nonlinear rheological behavior of asphalt, the large amplitude oscillation shear(LAOS) test was introduced, and the F...To reveal the effects of environmental and loading conditions, as well as asphalt properties on the nonlinear rheological behavior of asphalt, the large amplitude oscillation shear(LAOS) test was introduced, and the Fourier transform rheology, Lissajous curve method, and the LAOS fatigue test have been applied to investigate the nonlinear rheological behavior of asphalt binders. The research results indicate that a decrease in temperature, an increase in shear frequency and strain level, the introduction of polymer modifiers, and the aging effect of asphalt can significantly increase the nonlinearity of asphalt, manifested by the higher relative magnitude of the third harmonic and zero-strain nonlinear coefficient. For the two polymer modifiers selected in this study, the 4%polyurethane modifier exhibits a higher nonlinear lifting effect than the 4% styrene-butadiene-styrene(SBS). The impact of long-term aging on nonlinear viscoelasticity is observably greater than that of short-term aging. The zero-strain nonlinear coefficient estimated based on the average value method can accurately characterize the nonlinear viscoelasticity of asphalt, which can serve as an effective supplement to the relative magnitude of the third harmonic. All asphalts exhibit shear thinning behavior under the test temperature of 24℃, and the decrease in test temperature, the increase in shear rate and strain level, the introduction of modifiers, and the aging effect of asphalt all exacerbate the shear thinning behavior of asphalt. In addition, the fatigue failure process of asphalt materials is accompanied by an increasing degree of nonlinearity.展开更多
Stock return prediction has been in the spotlight because it involves numerous factors.Improving the accuracy of stock return prediction and quantifying the impact of individual factors on forecasting remain challengi...Stock return prediction has been in the spotlight because it involves numerous factors.Improving the accuracy of stock return prediction and quantifying the impact of individual factors on forecasting remain challenging tasks.Motivated by these challenges,we propose a novel forecasting method that entails proxy variables of category factors and the random forest technique.This new method aims to quantify the information and importance of category factors,thereby enhancing the predictability of stock returns.Specifically,we categorize a large set of return predictors into several category factors.We then utilize the importance of the original variables to construct proxy variables for these category factors.Subsequently,we use the proxy variables to build a random forest model for predicting stock returns.Our empirical analysis results demonstrate that the proposed method effectively quantifies the importance of both the original factors and category factors.Furthermore,we find that the fundamental information factor consistently ranks as the most crucial category factor for stock return forecasting.Additionally,the proposed method exhibits a more robust and prominent prediction performance than competing models such as single-category-factor-based random forest models,dimension-reduction,and forecast-combination methods.Most importantly,the proposed method produces forecast results that can assist investors with understanding stock market dynamics and facilitate higher investment returns.展开更多
Human pose estimation has received much attention from the research community because of its wide range of applications.However,current research for pose estimation is usually complex and computationally intensive,esp...Human pose estimation has received much attention from the research community because of its wide range of applications.However,current research for pose estimation is usually complex and computationally intensive,especially the feature loss problems in the feature fusion process.To address the above problems,we propose a lightweight human pose estimation network based on multi-attention mechanism(LMANet).In our method,network parameters can be significantly reduced by lightweighting the bottleneck blocks with depth-wise separable convolution on the high-resolution networks.After that,we also introduce a multi-attention mechanism to improve the model prediction accuracy,and the channel attention module is added in the initial stage of the network to enhance the local cross-channel information interaction.More importantly,we inject spatial crossawareness module in the multi-scale feature fusion stage to reduce the spatial information loss during feature extraction.Extensive experiments on COCO2017 dataset and MPII dataset show that LMANet can guarantee a higher prediction accuracy with fewer network parameters and computational effort.Compared with the highresolution network HRNet,the number of parameters and the computational complexity of the network are reduced by 67%and 73%,respectively.展开更多
This study examines a comprehensive set of 30 cross-sectional anomalies in the Chinese A-share market to investigate whether incorporating investor sentiment as conditioning information enhances the explanatory power ...This study examines a comprehensive set of 30 cross-sectional anomalies in the Chinese A-share market to investigate whether incorporating investor sentiment as conditioning information enhances the explanatory power of asset pricing models.Utilizing a long–short portfolio strategy and Fama–MacBeth cross-sectional regression,we find that trading-based anomalies outnumber accounting-based anomalies in the Chinese market.Our results demonstrate that conditional models significantly outperform their unconditional counterparts.Notably,investor sentiment is crucial for capturing the size anomaly when excluding observations from the COVID-19 pandemic period.Additionally,it substantially improves the ability of conditional Fama–French three-factor models to capture individual anomalies and enhances the return–prediction accuracy of conditional CAPMs.We suggest further investigating high-frequency investor sentiment-based conditional models to anticipate stock price fluctuations during extraordinary public health events.展开更多
基金Under the auspices of National Key R&D Program of China(No.2019YFC0409104)。
文摘Water conservation initiatives satisfy the demand for water supply,electricity generation,irrigation,and flood control.While helping humanity,they have also altered the ecosystem of natural rivers,impacted river ecology,disrupted river continuity,and jeopardized the existence of aquatic creatures in rivers.Studying the impacts of dam construction on rivers can enhance our knowledge of river ecological and environmental concerns and help sustain the health of river ecosystems,thereby realizing the harmony between humans and water in both theoretical and practical aspects.This study used bibliometrics and constructed an author-keyword 2-mode matrix network using Co-Occurrence software to identify the hotspots and research trend in eco-hydrology of dammed rivers.We identified‘FLOW’‘SEDIMENT’‘QUALITY’and‘MODEL’as the research hotspots in the ecological impact of dammed rivers,and combined the related literatures,we highlight the research progress in the four directions.Then the research shortcomings and prospect were discussed,including strengthening the monitoring and analysis of critical ecological variables,enhancing the hydrological monitoring density for small rivers,strengthening the research of relationship between eutrophication and zooplankton,establishing multiscale approaches,and combining multi-sources information technologies to improve parameter accuracy in the model research.
基金Supported by the National Key Research and Development Program of China(2021YFB2012601)National Natural Science Foundation of China(12204109)+1 种基金Science and Technology Innovation Plan of Shanghai Science and Technology Commission(21JC1400200)Higher Education Indus⁃try Support Program of Gansu Province(2022CYZC-06)。
文摘Organic semiconductor materials have shown unique advantages in the development of optoelectronic devices due to their ease of preparation,low cost,lightweight,and flexibility.In this work,we explored the application of the organic semiconductor Y6-1O single crystal in photodetection devices.Firstly,Y6-1O single crystal material was prepared on a silicon substrate using solution droplet casting method.The optical properties of Y6-1O material were characterized by polarized optical microscopy,fluorescence spectroscopy,etc.,confirming its highly single crystalline performance and emission properties in the near-infrared region.Phototransistors based on Y6-1O materials with different thicknesses were then fabricated and tested.It was found that the devices exhibited good visible to near-infrared photoresponse,with the maximum photoresponse in the near-infrared region at 785 nm.The photocurrent on/off ratio reaches 10^(2),and photoresponsivity reaches 16 mA/W.It was also found that the spectral response of the device could be regulated by gate voltage as well as the material thickness,providing important conditions for optimizing the performance of near-infrared photodetectors.This study not only demonstrates the excellent performance of organic phototransistors based on Y6-1O single crystal material in near-infrared detection but also provides new ideas and directions for the future development of infrared detectors.
基金Our research was funded by the Sichuan Key Provincial Research Base of Intelligent Tourism(No.ZHZJ23-02)supported by the Scientific Research and Innovation Team Program of Sichuan University of Science and Engineering(No.SUSE652A006)+1 种基金Additional support was provided by the National Cultural and Tourism Science and Technology Innovation Research andDevelopment Project(No.202417)the Lantern Culture and Crafts Innovation Key Laboratory Project of the Sichuan ProvincialDepartment of Culture and Tourism(No.SCWLCD-A02).
文摘The Quadric Error Metrics(QEM)algorithm is a widely used method for mesh simplification;however,it often struggles to preserve high-frequency geometric details,leading to the loss of salient features.To address this limitation,we propose the Salient Feature Sampling Points-based QEM(SFSP-QEM)—also referred to as the Deep Learning-Based Salient Feature-Preserving Algorithm for Mesh Simplification—which incorporates a Salient Feature-Preserving Point Sampler(SFSP).This module leverages deep learning techniques to prioritize the preservation of key geometric features during simplification.Experimental results demonstrate that SFSP-QEM significantly outperforms traditional QEM in preserving geometric details.Specifically,for general models from the Stanford 3D Scanning Repository,which represent typical mesh structures used in mesh simplification benchmarks,the Hausdorff distance of simplified models using SFSP-QEM is reduced by an average of 46.58% compared to those simplified using traditional QEM.In customized models such as the Zigong Lantern used in cultural heritage preservation,SFSP-QEM achieves an average reduction of 28.99% in Hausdorff distance.Moreover,the running time of this method is only 6%longer than that of traditional QEM while significantly improving the preservation of geometric details.These results demonstrate that SFSP-QEMis particularly effective for applications requiring high-fidelity simplification while retaining critical features.
基金supported by the National Science Foundation of China(Grant Nos.52068049 and 51908266)the Science Fund for Distinguished Young Scholars of Gansu Province(No.21JR7RA267)Hongliu Outstanding Young Talents Program of Lanzhou University of Technology.
文摘Addressing the current challenges in transforming pixel displacement into physical displacement in visual monitoring technologies,as well as the inability to achieve precise full-field monitoring,this paper proposes a method for identifying the structural dynamic characteristics of wind turbines based on visual monitoring data fusion.Firstly,the Lucas-Kanade Tomasi(LKT)optical flow method and a multi-region of interest(ROI)monitoring structure are employed to track pixel displacements,which are subsequently subjected to band pass filtering and resampling operations.Secondly,the actual displacement time history is derived through double integration of the acquired acceleration data and subsequent band pass filtering.The scale factor is obtained by applying the least squares method to compare the visual displacement with the displacement derived from double integration of the acceleration data.Based on this,the multi-point displacement time histories under physical coordinates are obtained using the vision data and the scale factor.Subsequently,when visual monitoring of displacements becomes impossible due to issues such as image blurring or lens occlusion,the structural vibration equation and boundary condition constraints,among other key parameters,are employed to predict the displacements at unknown monitoring points,thereby enabling full-field displacement monitoring and dynamic characteristic testing of the structure.Finally,a small-scale shaking table test was conducted on a simulated wind turbine structure undergoing shutdown to validate the dynamic characteristics of the proposed method through test verification.The research results indicate that the proposed method achieves a time-domain error within the submillimeter range and a frequency-domain accuracy of over 99%,effectively monitoring the full-field structural dynamic characteristics of wind turbines and providing a basis for the condition assessment of wind turbine structures.
基金financed by the grants from Dunhuang Medical Literature Compilation and Application Research Center Project of Gansu Provincial Key Research Base for Humanities and Social Sciences(No.DHYXJD2025)National Social Science Fund General Project(No.22BYY038).
文摘Dunhuang medicine is an important part of Dunhuang studies,boasting rich and comprehensive connotations.It records more than 1,000 Dunhuang medical prescriptions,involving internal,external,gynecological,pediatric,Buddhist,and Daoist medicine,and has demonstrated good clinical effects.However,the mechanism of action of relevant Dunhuang medical prescriptions is still unclear,existing research lacks systematic review and summarization,which has limited their further development.At the same time,the inheritance,innovation,and transformation of Dunhuang medicine are critical issues for the development of Dunhuang medicine,which has important guiding significance for the future development of Dunhuang medicine.Therefore,this study systematically summarizes the experimental research progress of Dunhuang medical prescriptions[except for those contained in Fu Xing Jue Zang Fu Yong Yao Fa Yao(《辅行诀脏腑用药法要》The Guideline to Use Medicines for Zang-fu)],and seven such prescriptions were selected based on three criteria:well-preserved texts,no prior transmission to the outside world,and having extensive research and clinical application over the past decade.The findings indicate that this type of prescription is applicable to a broad spectrum of diseases and has a promising application prospect in health preservation and disease prevention,as it exerts therapeutic effects through multiple targets and pathways.Based on this,specific strategies for the transformation of Dunhuang characteristic prescriptions were proposed from three aspects:inheritance,innovative development,and transformation strategies,aiming to provide insights for the future development of Dunhuang medical prescriptions.
基金supported by Biological Breeding of Early Maturing and Disease Resistant Cotton Varieties (NO.2023ZD04041)the Project of China Agriculture Research System (Grant No. CARS-15-06)+2 种基金Natural Science Foundation of Henan Province (Grant No. 232300421041 and 222300420382)National Natural Science Foundation of China (Grant No. U21 A20213)the Central Public-interest Scientific Institution Basal Research Fund (Grant No. 1610162023017 and 1610162023028)。
文摘Background Cotton is an important crop providing the most natural fibers all over the world. The cotton genomics community has utilized whole genome sequencing data to construct an elite gene pool in which functional genes are related to agronomic traits. However, the functional validation of these genes is hindered by time-consuming and inefficient genetic transformation methods. Thus, establishing a transient transformation system of high efficiency is necessary for cotton genomics.Results To improve the efficiency of transient transformation, we used the protoplasts isolated from the etiolated cotyledon as recipient. The enzymatic digestion buffer comprised 1.5%(w/v) cellulase, 0.75%(w/v) macerozyme, and 1% hemicellulase, osmotically buffered with 0.4 mol·L^(-1) mannitol. After 5 h of dark incubation at 25℃, uniform cotton protoplasts were successfully isolated with a yield of 4.6 × 10^(6) protoplasts per gram(fresh weight) and 95% viability. We incubated 100 μL protoplasts(2.5 × 10^(5)·m L^(-1)) with 15 μg plasmid in the solution of 0.4 mol·L^(-1) mannitol and 40% PEG 4000 for 15 min, ultimately achieving an optimal transient transfection efficiency of 71.47%.Conclusions This transient system demonstrated effective utility in cellular biology research through successful applications in subcellular localization analyses, bimolecular fluorescence complementation(Bi FC) verification, and prime editing vector validation. Through systematic optimization, we established an efficient and expedited protoplast-based transient transformation system and successfully applied this platform to cotton functional genomics studies.
基金supported by the National Key Research and Development Program of China(2023YFB2603500).
文摘To reveal the effects of environmental and loading conditions, as well as asphalt properties on the nonlinear rheological behavior of asphalt, the large amplitude oscillation shear(LAOS) test was introduced, and the Fourier transform rheology, Lissajous curve method, and the LAOS fatigue test have been applied to investigate the nonlinear rheological behavior of asphalt binders. The research results indicate that a decrease in temperature, an increase in shear frequency and strain level, the introduction of polymer modifiers, and the aging effect of asphalt can significantly increase the nonlinearity of asphalt, manifested by the higher relative magnitude of the third harmonic and zero-strain nonlinear coefficient. For the two polymer modifiers selected in this study, the 4%polyurethane modifier exhibits a higher nonlinear lifting effect than the 4% styrene-butadiene-styrene(SBS). The impact of long-term aging on nonlinear viscoelasticity is observably greater than that of short-term aging. The zero-strain nonlinear coefficient estimated based on the average value method can accurately characterize the nonlinear viscoelasticity of asphalt, which can serve as an effective supplement to the relative magnitude of the third harmonic. All asphalts exhibit shear thinning behavior under the test temperature of 24℃, and the decrease in test temperature, the increase in shear rate and strain level, the introduction of modifiers, and the aging effect of asphalt all exacerbate the shear thinning behavior of asphalt. In addition, the fatigue failure process of asphalt materials is accompanied by an increasing degree of nonlinearity.
基金supported by the National Natural Science Foundation of China(Grant No.72403117,U1901223,72271095)the Natural Science Foundation of Jiangsu Province,China(Grant No.BK20241435)+1 种基金Guangzhou Philosophy and Social Science Project(Grant No.x2gsN5180360)the Fundamental Research Funds for the Central Universities(Grant No.30923011034).
文摘Stock return prediction has been in the spotlight because it involves numerous factors.Improving the accuracy of stock return prediction and quantifying the impact of individual factors on forecasting remain challenging tasks.Motivated by these challenges,we propose a novel forecasting method that entails proxy variables of category factors and the random forest technique.This new method aims to quantify the information and importance of category factors,thereby enhancing the predictability of stock returns.Specifically,we categorize a large set of return predictors into several category factors.We then utilize the importance of the original variables to construct proxy variables for these category factors.Subsequently,we use the proxy variables to build a random forest model for predicting stock returns.Our empirical analysis results demonstrate that the proposed method effectively quantifies the importance of both the original factors and category factors.Furthermore,we find that the fundamental information factor consistently ranks as the most crucial category factor for stock return forecasting.Additionally,the proposed method exhibits a more robust and prominent prediction performance than competing models such as single-category-factor-based random forest models,dimension-reduction,and forecast-combination methods.Most importantly,the proposed method produces forecast results that can assist investors with understanding stock market dynamics and facilitate higher investment returns.
基金the National Natural Science Foundation of China(Nos.61775139,62072126,61772164,and 61872242)。
文摘Human pose estimation has received much attention from the research community because of its wide range of applications.However,current research for pose estimation is usually complex and computationally intensive,especially the feature loss problems in the feature fusion process.To address the above problems,we propose a lightweight human pose estimation network based on multi-attention mechanism(LMANet).In our method,network parameters can be significantly reduced by lightweighting the bottleneck blocks with depth-wise separable convolution on the high-resolution networks.After that,we also introduce a multi-attention mechanism to improve the model prediction accuracy,and the channel attention module is added in the initial stage of the network to enhance the local cross-channel information interaction.More importantly,we inject spatial crossawareness module in the multi-scale feature fusion stage to reduce the spatial information loss during feature extraction.Extensive experiments on COCO2017 dataset and MPII dataset show that LMANet can guarantee a higher prediction accuracy with fewer network parameters and computational effort.Compared with the highresolution network HRNet,the number of parameters and the computational complexity of the network are reduced by 67%and 73%,respectively.
基金financially supported by:National Natural Science Foundation of China(72261002,72141304)Youth Foundation for Humanities and Social Sciences Research of the Ministry of Education(22YJC790190)+1 种基金National Key Research and Development Program of China(2022YFC3303304)Student Research Program of Guizhou University of Finance and Economics(2022ZXS).
文摘This study examines a comprehensive set of 30 cross-sectional anomalies in the Chinese A-share market to investigate whether incorporating investor sentiment as conditioning information enhances the explanatory power of asset pricing models.Utilizing a long–short portfolio strategy and Fama–MacBeth cross-sectional regression,we find that trading-based anomalies outnumber accounting-based anomalies in the Chinese market.Our results demonstrate that conditional models significantly outperform their unconditional counterparts.Notably,investor sentiment is crucial for capturing the size anomaly when excluding observations from the COVID-19 pandemic period.Additionally,it substantially improves the ability of conditional Fama–French three-factor models to capture individual anomalies and enhances the return–prediction accuracy of conditional CAPMs.We suggest further investigating high-frequency investor sentiment-based conditional models to anticipate stock price fluctuations during extraordinary public health events.
