36,490 results on '"WU Yu"'
Search Results
2. Analysis of 39 Forensic Clinical Sex Identification Cases after Sex Reassignment Surgery
- Author
-
WU Yu-jiao, ZHOU Yue-qin, SHEN Yi-wen, and JIANG Jie-qing
- Subjects
forensic medicine ,sex identification ,transsexualism ,sex reassignment surgery ,cases analysis ,Medicine - Abstract
ObjectiveTo analyze the characteristics of forensic clinical sex identification cases after sex reassignment surgery for transsexualism, discuss the main point of forensic clinical sex identification in China based on domestic and foreign sex change policies and related cases, so as to standardize relevant identification.MethodsA total of 39 cases of forensic clinical sex identification were collected from Center for Forensic Science, Shanghai Medical College of Fudan University in recent years, and the ratio of sex, identification time, the location of surgery, secondary sex characteristic and other information were statistically analyzed.ResultsTransgender woman accounted for 97.4% of the cases, and 94.9% of the cases were operated in Thailand. All cases were identified after the sex reassignment surgery, and at least one of the secondary sex characteristics was consistent with the postoperative primary sexual characters. There were only 2 cases in which auxiliary examination was applied.ConclusionForensic clinical sex identification should be based on the principle that the primary and secondary sex characteristics are consistent with the changed sex.
- Published
- 2024
- Full Text
- View/download PDF
3. Research Status of Embryonic Developmental Toxicity and Biological Monitoring of Exposure to Traditional and Emerging Mycotoxins during Pregnancy
- Author
-
HAN Xiao-min, YE Jin, JI Jian, XU Wen-jing, WU Yu, YE Yong-li, Rudolf KRSKA, SUN Xiu-lan, and BAI Li
- Subjects
pregnancy ,traditional and emerging mycotoxin ,embryo developmental toxicity ,biological monitoring ,Food processing and manufacture ,TP368-456 ,Nutrition. Foods and food supply ,TX341-641 - Abstract
Mycotoxins are toxic secondary metabolites produced by certain fungal species under specific conditions. Research has found that exposure to various fungal toxins can cause embryonic developmental toxicity, such as embryo arrest, abortion, and intrauterine fetal developmental delay. There are no systematic reports on the impact of exposure to fungal toxins during pregnancy on the toxicity of intrauterine fetal development and biological monitoring. Therefore, in order to prevent and reduce the harm of fungal toxins during pregnancy, this paper summarizes the research status of embryonic developmental toxicity, as well as the research status of biological monitoring technologies based on biomarkers such as exposure biomarkers and biomakers of effect, which are caused by traditional fungal toxins such as aflatoxin, deoxynivalenol, zearalenone, and emerging fungal toxins such as Alternaria toxins, beauverin, and enniatins. The aim of the paper is for construction of the healthy China.
- Published
- 2024
- Full Text
- View/download PDF
4. Impact of individual income inequality on the sense of gain of farmers in poverty alleviation relocation: Mediation effect based on social integration
- Author
-
WU Yu, LI Jie
- Subjects
poverty alleviation relocation (par) ,sense of gain ,individual income inequality ,relative deprivation ,social integration ,shaanxi province ,Environmental sciences ,GE1-350 ,Biology (General) ,QH301-705.5 - Abstract
[Objective] Registered impoverished farmers’ sense of gain is crucial for determining the achievements of poverty alleviation relocation. This study aims to examine the impact of individual income inequality on sense of gain of farmers relocated for poverty alleviation and its mechanism after poverty alleviation relocation. [Methods] Based on the relative deprivation theory and data from 1202 questionnaires of farmers relocated for poverty alleviation collected in four cities and eight counties (districts) in Shaanxi Province in 2021, this study explores the relationship between individual income inequality, relocated farmers’ sense of gain, and social integration by using an ordered-probit model and the mediation effect model. [Results] (1) On the whole, the surveyed farmers think the gains outweigh the losses from the poverty alleviation relocation. However, individual income inequality has a significant negative impact on relocated farmers’ sense of gain. (2) The mediation mechanism analysis reveals that individual income inequality influences sense of gain through relocated farmers’ social integration, completely through economic integration while partly through social connections and community participation. Psychological integration has a suppression effect. (3) Heterogeneity analysis shows that the above effects of individual income inequality are weaker or insignificant among relocated farmers whose family obtain educational or medical and health supports. [Conclusion] The findings indicate that the government should not only implement employment policies targeted at low-income relocated households, but also help them expand their social networks. Meanwhile, it is necessary to improve the basic public service system in the destination areas.
- Published
- 2024
- Full Text
- View/download PDF
5. Research on Dynamic Characteristics of the Engine Front-end Accessory Drive System
- Author
-
Zheng Guangze, Yan Pengyuan, Wu Yu, and Li Xiaofeng
- Subjects
Multi-wedge belt ,Belt drive system ,Galerkin method ,Implicit multi-step method ,Dynamic characteristics ,Mechanical engineering and machinery ,TJ1-1570 - Abstract
Considering the effects of time-varying load, time-varying excitation, longitudinal stiffness, radial stiffness and bending stiffness of each pulley, the dynamic analysis model of the inner and outer engine front-end accessory drive system is established by Simdrive software. The static characteristics of the system are evaluated. Galerkin method is used to discretize the system space, the implicit multi-step method is used to solve the dynamic characteristics of the system under acceleration conditions, and the influencing factors of the accessory pulley speed fluctuation, tension fluctuation and belt-pulley slip characteristics are obtained. Based on the analysis results, an improvement plan is proposed to appropriately increase the initial tension and the crankshaft wheel wrap angle, and add an automatic tensioner at the No.2 idler gear. The comparison analysis shows that the dynamic characteristics of the engine front-end accessory drive system are improved effectively when adding automatic tensioner.
- Published
- 2024
- Full Text
- View/download PDF
6. Progress in research and applications of additively manufactured nickel-based superalloy in aero-engines and gas turbines
- Author
-
WU Yu, CHEN Bingqing, LIU Wei, HUANG Shuai, SUN Bingbing, ZHANG Xuejun, CHEN Pei, and HUANG Chen
- Subjects
additive manufacturing(am) ,nickel-based superalloy ,aero-engine ,gas turbine ,microstructure ,tensile property ,Motor vehicles. Aeronautics. Astronautics ,TL1-4050 - Abstract
Nickel-based superalloy is an essential material to prepare hot-end components in aero-engines and gas turbines,due to its excellent mechanical properties under high temperature. Additive manufacturing(AM) is one of the most important techniques to fabricate superalloy components with complex geometry. In this paper,the research progress of microstructure and defects of AMed superalloy is reviewed. Based on the existing literature,tensile properties of GH3536,GH3625 and GH4169 are summarized. Typical applications of AMed superalloy components in aero-engines and gas turbines are presented. Finally,for the problems in existing investigations,it is suggested that the future research can focus on materials design,heat treatment/hot isostatic pressure process optimization,single crystal preparation,real-time monitoring technique development and internal surface treatment technique innovation.
- Published
- 2024
- Full Text
- View/download PDF
7. Effect of ultra-high-speed laser melting deposition process on performance of Al-Mg-Sc high-strength aluminum alloy
- Author
-
WEN Pin, DONG Ziliang, WU Yu, DONG Dingyang, LUO Xing, DU Borui, XU Peixin, and CHEN Fei
- Subjects
ultra-high speed laser melting deposition ,high strength aluminum alloy ,mechanical properties ,numerical simulation ,Motor vehicles. Aeronautics. Astronautics ,TL1-4050 - Abstract
In order to clarify the microstructure of Al-Mg-Sc high-strength aluminum alloy prepared by ultra-high speed laser melting deposition,and study the relationship between structure and properties,Al-Mg-Sc high-strength aluminum alloy was prepared by using 7075 aluminum alloy as matrix and self-developed equipment(LDF3000-40 laser melting deposition machine). The effect of laser scanning rate on the microstructure and tensile properties of the materials was investigated. The results show that there are no obvious defects such as pores and cracks in the samples deposited by ultra-high speed laser melting,but contain a few small keyholes. The samples are composed of fine α-Al equiaxial crystals and dispersed Al3(Sc,Zr)particles. The effect of scanning rate on mechanical properties is further studied by numerical simulation. It is found that the faster scanning rate of laser in the range of 0.1-1 m/s can reduce the accumulation of powder materials and porosity of the surface of the deposition layer,thus the mechanical properties can be improved. The maximum tensile strength is 303 MPa and elongation at break is 22.5%.
- Published
- 2024
- Full Text
- View/download PDF
8. Application and research progress of recombinase isothermal amplification technology in the analysis of animal and plant components
- Author
-
SHI Yunxue, WANG Xin, WANG Jun, YAO Hua, and WU Yu
- Subjects
recombinase amplification technology ,animal and plant sources ,rapid detection ,Food processing and manufacture ,TP368-456 ,Nutrition. Foods and food supply ,TX341-641 - Abstract
Recently, the analysis of animal- and plant-derived components, mainly based on the adulteration and smuggling of meat products and genetically modified crops, has been a key focus in the field of food safety. Therefore, the requirement for analytical methods has been increasingly emphasized. Recombinase isothermal amplification is a novel and promising detection technique that currently includes recombinase polymerase amplification (RPA) and recombinase-aided amplification (RAA). This technique has been widely used in animal- and plant-derived component analysis and other detection and analysis fields owing to its speed, convenience, and high sensitivity and specificity. This paper summarizes relevant research conducted locally and abroad in recent years and concludes with an analysis of various detection methods. It also generalizes the characteristics and applications of RPA/RAA technology in four aspects: primer-probe design, external factors, reaction devices, and detection methods, and lays out the prospects for RPA/RAA development trends. Finally, this study also provides a reference for improving the analysis and detection methods of animal- and plant-derived components and promoting RPA/RAA technology.
- Published
- 2024
- Full Text
- View/download PDF
9. Accurate Determination of Aflatoxin in Grains and Oils use Liquid Chromatography Mass Spectrometry Purified by Automatic Solid Phase Extraction Instrument and Quantified by Fully 13C-Labelled Internal Standards
- Author
-
YE Jin, LIU Tong-tong, LUO Li-juan, WU Yu, LI Li, and WANG Song-xue
- Subjects
automatic solid phase extraction instrument ,ultra-high performance liquid chromatography- quadrupole orbitrap high-resolution mass spectrometric ,aflatoxin ,immunoaffinity column ,fully 13c isotope labeled internal standard calibration method ,Food processing and manufacture ,TP368-456 ,Nutrition. Foods and food supply ,TX341-641 - Abstract
An UPLC Quadrupole Orbitrap high resolution mass Spectrometry method for accurate quantification of aflatoxins in grains and oils was established, which were purified by automatic immunoaffinity column. The sample was extracted with 84/16(V/V) acetonitrile/water and centrifuged. The supernatant was diluted 10 times with 0.1% PBST. Aflatoxin immunoaffinity column without background and recovery greater than 90% was chosen for purification automatically. UPLC-Quadrupole Orbitrap HRMS was used for quantitative analysis. Results showed that recoveries for aflatoxins were 84.7%~104.6%. The intra-day relative standard deviations were 1.7%~9.4%, and the intra-day precision were 0.7%~8.4%. The developed method can minimize the workload, improve work efficiency and avoid the result deviation caused by operation manually. It is suitable for the accurate determination of aflatoxins in grains and oils, and meet the requirement of the daily test and monitoring aflatoxins.
- Published
- 2023
- Full Text
- View/download PDF
10. Comparison of the efficacy and safety between rivaroxaban and dabigatran in the treatment of acute portal vein thrombosis in cirrhosis
- Author
-
Haonan Zhou, Mingdong Wu, Shixiong Yu, Han Xia, Wu Yu, Kai Huang, and Yikuan Chen
- Subjects
Acute portal vein thrombosis ,Rivaroxaban ,Dabigatran ,Cirrhosis ,Diseases of the digestive system. Gastroenterology ,RC799-869 - Abstract
Abstract Background New oral anticoagulants (NOACs) have been becoming prevalent in recent years and are increasingly used in the treatment of port vein thrombosis. The difference of the efficacy and safety between rivaroxaban and dabigatran remains unclear in the treatment of cirrhotic patients with acute portal vein thrombosis (PVT). Methods This retrospective study included all consecutive cirrhotic patients with acute portal vein thrombosis in our institute from January 2020 to December 2021. The patients received oral anticoagulation with rivaroxaban or dabigatran. The demographic, clinical, and imaging data of patients were collected. The diagnosis of acute PVT was confirmed by imaging examinations. The severity of liver cirrhosis was assessed using Child–Pugh score and Model for End-Stage Liver Disease (MELD) score. Outcomes included recanalization (complete, partial, and persistent occlusion), liver function, bleedings, and survival. The log-rank test was used to compare Kaplan–Meier distributions of time-to-event outcomes. The Cox proportional hazards model was used to calculate hazard ratios (HRs) with 95% confidence intervals (CIs). Results A total of 94 patients were included, 52 patients (55%) received rivaroxaban and 42 (45%) with dabigatran. The complete and partial recanalization of PVT was observed in 41 patients. There was no significant difference in complete recanalization, partial recanalization, and persistent occlusion between the two groups. With multivariate analysis, D-dimer (HR 1.165, 95% CI 1.036–1.311, p = 0.011) was independent predictors of complete recanalization. The Child–Pugh score (p = 0.001) was significantly improved in both two groups after anticoagulation, respectively. However, there was no difference between the two groups. The probability of survival was 94%, 95% in the rivaroxaban and dabigatran groups (log-rank p = 0.830). Major bleedings were reported in 3 patients (6%) in rivaroxaban group and 1 patient (2%) in dabigatran group (p = 0.646). Six patients (12%) in rivaroxaban group experienced minor bleeding, and five (12%) from dabigatran group (p = 0.691). Conclusions The efficacy and safety were comparable between rivaroxaban and dabigatran in the treatment of cirrhotic patients with acute portal vein thrombosis. And D-dimer can contribute to the prediction of PVT recanalization in cirrhotic patients.
- Published
- 2023
- Full Text
- View/download PDF
11. Decoupling of soil carbon mineralization and microbial community composition across a climate gradient on the Tibetan Plateau
- Author
-
Jingyang Zheng, Xiali Mao, Kees Jan van Groenigen, Shuai Zhang, Mingming Wang, Xiaowei Guo, Wu Yu, Lun Luo, Jinfeng Chang, Zhou Shi, and Zhongkui Luo
- Subjects
Bacteria ,Carbon fractions ,Fungi ,Microbial diversity ,r/K strategists ,Soil organic carbon ,Science - Abstract
Soil microbes drive soil organic carbon (SOC) mineralization. Because microbial groups differ in metabolic efficiency and respond differently to temperature variation, it is reasonable to expect a close association of SOC mineralization and its temperature sensitivity (Q10 which is defined as the factor of the change of soil carbon mineralization induced by 10 °C temperature increase) with microbial community diversity and composition. However, these relations have rarely been tested. Here, we conducted an incubation experiment to assess the temperature responses of microbial α diversity and the relative abundance of microbial r- and K-strategists in soils from a wide range of ecosystems across a climate gradient in the southeast Tibet. The results indicated that the instantaneous α diversity and the relative abundance of r- and K-strategists are significantly (P
- Published
- 2024
- Full Text
- View/download PDF
12. The 980 nm Pump Laser with Output Power of 400 mW
- Author
-
LIU Jun, LI Bo-zhong, CHENG Fang, LI Zi-fan, TIAN Zhao-yu, QIAN Sheng-qi, WU Yu, QU Xue-yang, and LIU Rui
- Subjects
980 nm wavelength ,pump laser ,400 mW output power ,wavelength stability ,Applied optics. Photonics ,TA1501-1820 - Abstract
Due to the lack of domestic 980 nm pump laser design experience and production process, the key technology of high power 980 nm pumped laser to break through. In this paper, a 980 nm pump laser with 400 mW output power has been successfully developed by using a domestic solution, including the design and implementation of chips, gratings and coupling fibers. By using the all metallized package directly coupled with wedge-shaped fiber and selecting the double grating structure for wavelength locking, the wavelength stability of the 980 nm pumped laser with 400 mW output power in the operating temperature range of -50~75 ℃ can reach 0.01 nm/ ℃. Moreover the 980 nm pump laser has passed the 800 mA charge aging at 75 ℃ for more than 5 000 h. This has laid a foundation for the research of domestic pump lasers with high reliability.
- Published
- 2023
- Full Text
- View/download PDF
13. Real‐World clinical features and survival outcomes associated with primary gastrointestinal natural killer/T‐cell lymphoma from 1999 to 2020
- Author
-
Yang Chunli, Jiang Ming, Ma Ziyan, Ji Jie, Lv Shuli, Huang Jie, Wu Yu, Xu Caigang, and Zou Liqun
- Subjects
extranodal NK/T‐cell lymphoma ,gastrointestinal ,nomogram ,prognosis ,survival ,Neoplasms. Tumors. Oncology. Including cancer and carcinogens ,RC254-282 - Abstract
Abstract Background Primary gastrointestinal natural killer (NK)/T‐cell lymphoma (PGINKTL) is a rare T‐/NK‐cell lymphoma subtype, and the clinical features and survival outcomes remain largely unknown. Methods To summarize the clinical features and survival outcomes of PGINKTL, PGINKTL cases diagnosed at our hospital from May 1999 to December 2020 were reviewed; and the clinical data, information on treatment strategies, and survival were collected. Survival analysis was performed using the Kaplan–Meier method and multivariable Cox proportional hazards regression. We constructed a nomogram to visualize the survival prediction of PGINKTL. The discriminative ability and calibration of the nomogram for prediction were tested using the concordance index (C‐index) and calibration plots. Results The cohort included 81 cases, the median age was 36 years (range, 7–80 years), and the male‐to‐female ratio was 1.7:1. The most common clinical symptom at the time of diagnosis was abdominal pain (71.6%). The most common lesion site was the colon (59.3%). During a median follow‐up period of 37.7 months, the median overall survival (OS) time of 81 patients was 4.0 months (95% confidence interval [CI], 3.1–4.9 months), and the 2‐year OS rate was 30.7% (95% CI, 20.3%–40.1%). The multivariate analyses indicated that patients with an Eastern Cooperative Oncology Group (ECOG) performance status (PS) score ≥2, serum lactic dehydrogenase (LDH) level ≥ the upper limit normal (ULN), and perforation had worse OS. We used these data to establish a nomogram to predict survival for PGINKTL. The nomogram displayed good accuracy, with a C‐index of 0.726. Conclusion The clinical features and poor outcomes of PGINKTL, which is a rare and fatal lymphoma type, are presented. The proposed nomogram provides an individualized estimate of survival for these patients. In the future, the study focused on exploring a better treatment strategy to improve survival is required in PGINKTL.
- Published
- 2023
- Full Text
- View/download PDF
14. Spatial Variation Characteristics and Influencing Factors of Soil Organic Carbon in a Rocky Beach of Sejila Mountain, Tibetan Plateau
- Author
-
Xiaojiao Wang, Wandong Liang, Dan Wan, Wu Yu, and Hui Yang
- Subjects
soil organic carbon ,soluble organic carbon ,easily oxidized organic carbon ,influencing factors ,redundancy analysis ,Environmental sciences ,GE1-350 ,General. Including nature conservation, geographical distribution ,QH1-199.5 - Abstract
[Objective] The characteristics of surface soil soluble organic carbon (DOC), readily oxidized organic carbon (ROC), carbon pool stability, and carbon storage changes, as well as influencing factors of soil organic carbon (SOC) and DOC at different elevations in Tibetan Plateau were analyzed in order to provide a scientific basis for subsequent research on the stability of SOC in cold regions with complex terrain. [Methods] Data were collected from a rocky beach of the Sejila Mountain. One-way ANOVA was used to analyze soil physical properties and DOC, ROC, and SOC content characteristics at altitudes ranging from 4 300 m to 4 700 m. Pearson correlation analysis and redundancy analysis were used to determine the relationship between DOC, ROC, and influencing factors. [Results] ① Soil bulk density and dry density were negatively correlated with altitude, and their maximum values were 1.56 g/cm3 and 1.12 g/cm3, respectively. Soil water content and porosity were positively correlated with altitude, and the maximum values were 74.22% and 58.70%, respectively. ② SOC content initially increased with increasing altitude and then decreased. DOC and ROC showed fluctuating trends with elevation. ③ DOC/SOC and ROC/SOC reached maximum values at 4 400 m. The average DOC and SOC values were 40.779 kg/m2 and 8.105 kg/m2, respectively. The stability of the soil carbon pool in the rocky beach area was relatively good. ④ DOC had no correlation with soil physical properties, but was correlated with average SOC. ROC was correlated with physical properties, SOC, and its related factors. Each factor was correlated with changes in DOC and ROC in the rocky beach area of the Sejila Mountains. Explanatory contribution rate follow the order of average DOC>soil porosity>ROC/SOC>soil water content>SOC>average SOC>DOC/SOC>soil dry density. [Conclusion] Soil DOC and ROC at different elevations were correlated, and average DOC, soil porosity, and ROC/SOC were important factors affecting the contents of DOC and ROC. There were close relationships among these factors.
- Published
- 2023
- Full Text
- View/download PDF
15. Study on main failure modes of multiple-holes structures based on branch-bound method
- Author
-
XIE Wei, ZHANG Yadan, MA Zehua, FENG Jinrong, and WU Yu
- Subjects
五孔结构 ,分枝-约界算法 ,寿命干涉模型 ,可靠性指标 ,主要失效模式 ,Motor vehicles. Aeronautics. Astronautics ,TL1-4050 - Abstract
Aiming at the typical five-hole structure on the aircraft skin, combined with the failure criterion of multi crack structure and considering the random factors affecting the residual life distribution of multiple crack structure, this paper applies the branch and bound algorithm to the screening of the main failure modes of typical five-hole structure, and gives the specific screening steps. The residual life of the structure is predicted by Franc3D crack propagation software, and the life interference model is established. Finally, the residual life distribution curve of the five-hole structure is formed to complete the screening of its main failure modes. The results show that the efficiency of using the algorithm to screen the main failure modes of the structure has been improved by 77.2%, and the error between the corresponding median life and that before screening is about 12%, indicating that the result of screening the main failure modes of the structure by the algorithm is more accurate. Compared with the existing test results, the median life predicted by the life distribution is more conservative and has certain reliability and accuracy. This method has certain reference significance for the evaluation of the safe life of this multi-holes structure.
- Published
- 2022
- Full Text
- View/download PDF
16. Divergent soil health responses to long-term inorganic and organic fertilization management on subtropical upland red soil in China
- Author
-
Si Shao-cheng, Wu Yu-cheng, Li Yuan, Yang Shuai, Peng Xiao-hong, and Luo Yong-ming
- Subjects
Long-term fertilization ,Soil fertility ,Heavy metal contamination ,Soil health ,Micronutrients ,Ecology ,QH540-549.5 - Abstract
To optimize soil health, land use, and soil management decisions need to be guided by an understanding of the underlying soil health. Although widely used fertilizers guarantee crop production, they ultimately damage agricultural soil environmental quality and threaten the overall food chain safety. Thus, an integrative assessment considering both environmental safety and soil fertility is required for the sustainable development of long-term fertilization systems. In this study, we conducted a comprehensive assessment of soil health using 20 soil physical, chemical, and biological indicators and assessed them according to four heavy metals—copper, zinc, cadmium, and lead. This experiment was conducted in an area where there is an ongoing, 23-year-long fertilization experiment in a typical red soil region of southern China, including four treatments of common inorganic chemical fertilizer (CF), CF mixed with micronutrients (CF + T), CF mixed with pig manure (PM), and PM mixed with micronutrients (PM + T). To assess the soil fertility quality, we selected a minimum data set consisting of soil water-stable aggregates mean weight diameter, available phosphorus, total nitrogen, microbial biomass carbon, and dehydrogenase activity. The results revealed that soil fertility quality followed the order of PM + T ≈ PM > CF + T > CF. We also used a modified potential ecological risk index to assess the soil environmental quality. This approach considered the acid-soluble and exchangeable fraction of heavy metals as the bioavailability component. Both PM and PM + T treated soils were significantly contaminated by heavy metals, with particularly large amounts of Cd contamination. The soil health index escalated from CF, CF + T, PM to PM + T, which indicated that the use of pig manure allowed for more fertile soil relative to standard chemical fertilization. Moreover, applying additional micronutrients also had a significant influence on soil fertility. Therefore, our study revealed that long-term fertilization using organic pig manure was beneficial to overall soil fertility; however, these benefits were coupled with significant limitations regarding heavy metal contamination. Collectively, this study provides a better supporting method for soil health management and long-term fertilization strategies.
- Published
- 2023
- Full Text
- View/download PDF
17. A Systematic Research on System Recovery Based on Improved Genetic Algorithm and Quotient Resilience Model Under Attack and Damage
- Author
-
Li Zhen, Tian Lu, Sun Chen Xu, Wu Yu Mei, Wang Dong Sheng, and Miao Hong
- Subjects
Resilience ,Explosion attack ,Recovery strategy ,Genetic algorithm ,Quotient resilience measurement model ,Electronic computers. Computer science ,QA75.5-76.95 - Abstract
Abstract The vulnerability of the current network has become an urgent problem to be solved. The focus of network protection should be shifted from traditional network protection to the direction of effective recovery even after being attacked and damaged, and then, the concept of resilience came into being. This paper selects physical explosion attacks to establish damaged network. An improved system resilience recovery strategy is established which considers task importance and time efficiency. Aiming at the initial population is too random, easy to mature and with poor solution, this paper improves genetic algorithm by new greedy model in population initialization and head-to-head mutation operator. Simulation shows that the improved genetic algorithm is better and more stable, the improved quotient model is more effective in system resilience recovery measured by index-E proposed in this paper.
- Published
- 2022
- Full Text
- View/download PDF
18. Effects of different environmental intervention durations on the intestinal mucosal barrier and the brain-gut axis in rats with colorectal cancer
- Author
-
Liu Dun, Chen Mei-Jing, Huang Si-Ting, Yu Xin-Yuan, and Wu Yu-Xuan
- Subjects
Medicine ,Science - Abstract
Abstract An enriched environment (EE) is a promising strategy for protecting the intestinal mucosal barrier and regulating the brain-gut axis, but the optimal EE intervention duration is unknown. Here, different EE intervention durations were applied to assess the optimal intervention duration in rats with colorectal cancer. We used a rat model of 1, 2-dimethylhydrazine-induced colorectal cancer. The rats were housed in an EE for 0, 2, 4, 8 weeks and 8-week blank group. The intestinal mucosa and serum TNF-α, IL-6, IL-10, ATP, CRF, and occludin levels and bacterial translocation (BT) were measured, and the intestinal mucosa morphology was evaluated. In 8 weeks, the effect of tumor on intestinal mucosal barrier was not obvious and the EE had a greater impact on it. Eight weeks of EE was more beneficial to the intestinal mucosal mechanical barrier than 2 or 4 weeks of intervention. A significant difference in BT was found between the 4- and 8-week groups. Overall, the analysis of inflammatory factor regulation revealed that the two blank groups exhibited the worst effect, and the intervention effect at 8 weeks was better than that at 2 and 4 weeks. CRF at 4 weeks was higher than that at 8-week blank group. The effect of 8-week intervention duration on the intestinal mucosal barrier was generally better than that of 2- and 4-week durations and intervention within 4 weeks can help to stabilize and promote the secretion of brain gut peptide, but the effect of different intervention durations on the brain-gut peptide levels was not obvious. In the future, we can further explore the molecular biological mechanism of the effect of different EE intervention durations on the intestinal mucosal barrier and analyze the effect of an EE on other brain-gut peptides.
- Published
- 2022
- Full Text
- View/download PDF
19. Research progress on the functional components of Grifola frondosa and their biological activities
- Author
-
XIONG Wen-yu, HE Jun-qiang, DAI Wan-zhen, WU Yu-yan, and LIU Bin
- Subjects
grifola frondosa ,functional components ,biological activity ,structure-activity relationship ,functional foods ,Food processing and manufacture ,TP368-456 - Abstract
The main functional components and biological activities of Grifola frondosa were reviewed, and the existing problems and future research directions were also discussed.
- Published
- 2022
- Full Text
- View/download PDF
20. Effect of folic acid, vitamin B12, and B6 supplementation on the risk of cardiovascular and cerebrovascular diseases: An updated meta-analysis of randomized controlled trials
- Author
-
Ye Man, Chen Xiaohong, Mao Sufang, Zhou Jing, Liu Minfen, and Wu Yu
- Subjects
vitamin b supplementation ,cardiovascular disease ,mortality ,meta-analysis ,Crystallography ,QD901-999 - Abstract
Numerous studies indicated that B vitamin supplementation can reduce cardiovascular risk; nonetheless, available proof reported from individual studies have not been consistent, so we performed an updated meta-analysis of randomized controlled trials (RCTs) to evaluate the relationship between B vitamin supplementation and cardiovascular outcomes.
- Published
- 2022
- Full Text
- View/download PDF
21. Neuroprotective effect and mechanism of idebenone on hippocampal damage induced by epileptic seizure
- Author
-
QIAO Shan, SU Yong-xin, ZHANG Ran-ran, WU Yu-jiao, WANG Ke-mo, and LIU Xue-wu
- Subjects
benzoquinones ,epilepsy ,hippocampus ,apoptosis ,superoxide dismutase ,glutathione peroxidase ,malondialdehyde ,disease models animal ,Neurology. Diseases of the nervous system ,RC346-429 - Abstract
Objective To investigate the protective effect of idebenone (IDBN) on hippocampal neuron injury in epileptic rats and its mechanism. Methods Forty-eight Wistar rats were randomly divided into normal control group, IDBN prevention group (prevention group), epilepsy group, IDBN 25, 50 and 100 mg groups. Behavioral changes of rats in different treatment groups before and after treatment with IDBN were observed. The activity of superoxide dismutase (SOD) and glutathione peroxidase (GSH-Px) and the content of malondialdehyde (MDA) in hippocampal were detected. The ultrastructural changes of hippocampal neurons and mitochondria were observed. Results The expression levels of SOD, GSH-Px and MDA in different treatment groups were significantly different (P = 0.000, for all). The activity of SOD and GSH-Px was the highest and the content of MDA was the lowest in IDBN 100 mg group. The activity of SOD and GSH-Px in IDBN 100 mg group were higher than those in epilepsy group (P = 0.000, 0.000), IDBN 25 mg group (P = 0.000, 0.000) and 50 mg group (P = 0.004, 0.005), while the content of MDA was lower than that in epilepsy group (P = 0.000), IDBN 25 mg group (P = 0.000) and 50 mg group (P = 0.002). Compared with normal control group and prevention group, hippocampal neurons in epilepsy group showed different degrees of damage, which was relieved after treatment with IDBN and gradually decreased with the increase of dose. In epilepsy group, the mitochondrial structure of hippocampal neurons was damaged obviously, with deformation and swelling, and some mitochondria were vacuolated. After treatment with IDBN, the damage was relieved, and the 100 mg group had the least damage. Conclusions IDBN can protect hippocampal neuron structure and function by inhibiting oxidative stress injury in epileptic rats.
- Published
- 2022
- Full Text
- View/download PDF
22. The Incubation Period of COVID-19 Caused by Different SARS-CoV-2 Variants
- Author
-
WU Yu, LIU Min
- Subjects
covid-19 ,sars-cov-2 ,variants ,infectious disease incubation period ,epidemiology ,review ,Medicine - Abstract
Since the beginning of the COVID-19 epidemic, the pathogen of COVID-19, SARS-CoV-2, has evolved and mutated continuously, producing variants with different enhanced transmission and virulence, such as Alpha (B.1.1.7) , Beta (B.1.351) , Gamma (P.1) , Delta (B.1.617.2) and Omicron (B.1.1.529) . An intensive study of the incubation period of COVID-19 caused by different SARS-CoV-2 variants will contribute to tracing the origin of COVID-19, determining the detention, quarantine and isolation time of close contacts, and timely improving measures for containing COVID-19. We reviewed the major studies on the incubation period of COVID-19 caused by wild-type strains and different variants of SARS-CoV-2, which estimated that the incubation period of COVID-19 caused by wild-type SARS-CoV-2 strains was 4-8 (median 5.5) days. And that for COVID-19 caused by Beta or Gamma variant was generally similar to that by wild-type strains, about 5 days. The incubation period of COVID-19 caused by Alpha, Delta and Omicron variants was shorter than that of other strains, which was 4, 4 and 3 days, respectively.
- Published
- 2022
- Full Text
- View/download PDF
23. Anomaly Detection Model Based on One-class Support Vector Machine Fused Deep Auto-encoder
- Author
-
WU Yu-kun, LI Wei, NI Min-ya, XU Zhi-cheng
- Subjects
deep auto-encoder ,one-class svm ,anomaly detection ,hybrid model ,Computer software ,QA76.75-76.765 ,Technology (General) ,T1-995 - Abstract
Large-scale high-dimensional unbalanced data handling is a major challenge in anomaly detection.One-class support vector machine(OCSVM) is very efficient at handling unbalanced data,but it is not suitable for large-scale high-dimensional dataset.Meanwhile,the kernel function of OCSVM also has an important influence on the detection performance.An anomaly detection model combining a deep auto-encoder and a one-class support vector machine is proposed.The deep auto-encoder is not only responsible for extracting features and dimensionality reduction,but also mapping an adaptive kernel function.As a whole,the model adopts the gradient descent method to carry out joint training and realizes end-to-end training.Experiment is conducted on four public datasets and compared with other anomaly detection methods.Experimental results show that the proposed model has better performance than single-kernel or multi-kernel one-class support vector machines and other models in terms of AUC and RECALL,and the proposed model is robust at different anomaly rate and has great advantages in time complexity.
- Published
- 2022
- Full Text
- View/download PDF
24. Analysis of Impact of Altitude on Vertical Distribution Pattern of Soil Organic Carbon in Mountainous Regions of Southeastern Tibet
- Author
-
Shalin HUANG, Jing LI, Wenji YANG, Tianmiao LIN, and Wu YU
- Subjects
alpine region ,sejila mountain ,altitude gradient ,physical property ,soil organic carbon ,vertical distribution ,Agriculture - Abstract
【Objective】The spatial heterogeneity of alpine ecosystems in southeastern Tibet in the vertical distribution zone was analyzed, and the distribution characteristics and influencing factors of soil organic carbon(SOC)reserves at different altitudes were studied.【Methods】Taking the Sejila Mountain, a typical mountain in southeastern Tibet in the alpine and semi-humid region, as the research object, sampling was carried out at an altitude of 3 400-4 600 m with an altitude gradient of 200 m, and the altitude was investigated by combining field investigation and indoor analysis. The impact of altitude on the vertical distribution characteristics of SOC and its correlation with other physical and chemical indicators.【Result】There was a great spatial heterogeneity among the soil physical indexes in the Sejila Mountain at an altitude of 3 400~4 600 m. The soil bulk density, total porosity, capillary porosity and non-capillary porosity varied 0.72-1.08 g/cm3, 51.76%-67.06%, 47.11%-58.03% and 9.03%-11.04%. SOC was mainly distributed in the 0-60 cm soil layer, and the SOC contents of different altitude gradients(1.57%-8.24%)showed a downward trend with the increase of soil depth, and the SOC content of the alpine meadow at an altitude of 4 618 m was the highest, which was 1.76-3.07 times that of other altitudes; SOC reserve was 117.14-220.63 t/hm2, among which the SOC reserve in the soil layer at an altitude of 4 255 m was significantly higher than that at other altitudes, equivalent to 2 times that at an altitude of 3 393 m; different altitudes significantly affected the SOC reserves in the 0-60 cm soil layer, showing a single peak as a whole curve change trend.【Conclusion】There are significant differences in the vertical distribution of SOC under different vegetation types in the Sejila Mountain, and the change of altitude gradient will directly or indirectly affects the accurate estimation of SOC, and then affects the spatial distribution and prediction of SOC.
- Published
- 2022
- Full Text
- View/download PDF
25. Epidemiologic Features and Containment of SARS-CoV-2 Omicron Variant
- Author
-
WU Yu, LIU Jue, LIU Min, LIANG Wannian
- Subjects
sars-cov-2 ,omicron variants of concern ,covid-19 ,epidemiology ,review ,Medicine - Abstract
Since the first case of the Omicron (B.1.1.529) variant discovered in South Africa was reported to the WHO on November 24, 2021, a total of 57 countries (regions) had reported Omicron cases as of December 8, 2021. Omicron has become the dominant strain in some African countries and is spreading rapidly. Although Omicron causes mild symptoms, with most cases being asymptomatic and mild, the rapid increase in the number of cases could put a heavy strain on global health systems. In addition, its source, transmission characteristics and vaccine resistance remain unclear, which brings great challenges to pandemic prevention and control in all countries (regions) . We reviewed the latest developments in etiological characteristics, mutation sources, transmission characteristics and possible mechanisms, pandemic status, vaccine protection effect and containment measures regarding Omicron, providing a reference for scientific containment of Omicron mutant.
- Published
- 2022
- Full Text
- View/download PDF
26. TV-Dialogue: Crafting Theme-Aware Video Dialogues with Immersive Interaction
- Author
-
Wang, Sai, Ma, Fan, Li, Xinyi, Fan, Hehe, and Wu, Yu
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Recent advancements in LLMs have accelerated the development of dialogue generation across text and images, yet video-based dialogue generation remains underexplored and presents unique challenges. In this paper, we introduce Theme-aware Video Dialogue Crafting (TVDC), a novel task aimed at generating new dialogues that align with video content and adhere to user-specified themes. We propose TV-Dialogue, a novel multi-modal agent framework that ensures both theme alignment (i.e., the dialogue revolves around the theme) and visual consistency (i.e., the dialogue matches the emotions and behaviors of characters in the video) by enabling real-time immersive interactions among video characters, thereby accurately understanding the video content and generating new dialogue that aligns with the given themes. To assess the generated dialogues, we present a multi-granularity evaluation benchmark with high accuracy, interpretability and reliability, demonstrating the effectiveness of TV-Dialogue on self-collected dataset over directly using existing LLMs. Extensive experiments reveal that TV-Dialogue can generate dialogues for videos of any length and any theme in a zero-shot manner without training. Our findings underscore the potential of TV-Dialogue for various applications, such as video re-creation, film dubbing and its use in downstream multimodal tasks.
- Published
- 2025
27. Learning to Help in Multi-Class Settings
- Author
-
Wu, Yu, Li, Yansong, Dong, Zeyu, Sathyavageeswaran, Nitya, and Sarwate, Anand D.
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence - Abstract
Deploying complex machine learning models on resource-constrained devices is challenging due to limited computational power, memory, and model retrainability. To address these limitations, a hybrid system can be established by augmenting the local model with a server-side model, where samples are selectively deferred by a rejector and then sent to the server for processing. The hybrid system enables efficient use of computational resources while minimizing the overhead associated with server usage. The recently proposed Learning to Help (L2H) model trains a server model given a fixed local (client) model, differing from the Learning to Defer (L2D) framework, which trains the client for a fixed (expert) server. In both L2D and L2H, the training includes learning a rejector at the client to determine when to query the server. In this work, we extend the L2H model from binary to multi-class classification problems and demonstrate its applicability in a number of different scenarios of practical interest in which access to the server may be limited by cost, availability, or policy. We derive a stage-switching surrogate loss function that is differentiable, convex, and consistent with the Bayes rule corresponding to the 0-1 loss for the L2H model. Experiments show that our proposed methods offer an efficient and practical solution for multi-class classification in resource-constrained environments., Comment: 30 pages, 7 figures, conference, ICLR 2025
- Published
- 2025
28. DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- Author
-
DeepSeek-AI, Guo, Daya, Yang, Dejian, Zhang, Haowei, Song, Junxiao, Zhang, Ruoyu, Xu, Runxin, Zhu, Qihao, Ma, Shirong, Wang, Peiyi, Bi, Xiao, Zhang, Xiaokang, Yu, Xingkai, Wu, Yu, Wu, Z. F., Gou, Zhibin, Shao, Zhihong, Li, Zhuoshu, Gao, Ziyi, Liu, Aixin, Xue, Bing, Wang, Bingxuan, Wu, Bochao, Feng, Bei, Lu, Chengda, Zhao, Chenggang, Deng, Chengqi, Zhang, Chenyu, Ruan, Chong, Dai, Damai, Chen, Deli, Ji, Dongjie, Li, Erhang, Lin, Fangyun, Dai, Fucong, Luo, Fuli, Hao, Guangbo, Chen, Guanting, Li, Guowei, Zhang, H., Bao, Han, Xu, Hanwei, Wang, Haocheng, Ding, Honghui, Xin, Huajian, Gao, Huazuo, Qu, Hui, Li, Hui, Guo, Jianzhong, Li, Jiashi, Wang, Jiawei, Chen, Jingchang, Yuan, Jingyang, Qiu, Junjie, Li, Junlong, Cai, J. L., Ni, Jiaqi, Liang, Jian, Chen, Jin, Dong, Kai, Hu, Kai, Gao, Kaige, Guan, Kang, Huang, Kexin, Yu, Kuai, Wang, Lean, Zhang, Lecong, Zhao, Liang, Wang, Litong, Zhang, Liyue, Xu, Lei, Xia, Leyi, Zhang, Mingchuan, Zhang, Minghua, Tang, Minghui, Li, Meng, Wang, Miaojun, Li, Mingming, Tian, Ning, Huang, Panpan, Zhang, Peng, Wang, Qiancheng, Chen, Qinyu, Du, Qiushi, Ge, Ruiqi, Zhang, Ruisong, Pan, Ruizhe, Wang, Runji, Chen, R. J., Jin, R. L., Chen, Ruyi, Lu, Shanghao, Zhou, Shangyan, Chen, Shanhuang, Ye, Shengfeng, Wang, Shiyu, Yu, Shuiping, Zhou, Shunfeng, Pan, Shuting, Li, S. S., Zhou, Shuang, Wu, Shaoqing, Yun, Tao, Pei, Tian, Sun, Tianyu, Wang, T., Zeng, Wangding, Zhao, Wanjia, Liu, Wen, Liang, Wenfeng, Gao, Wenjun, Yu, Wenqin, Zhang, Wentao, Xiao, W. L., An, Wei, Liu, Xiaodong, Wang, Xiaohan, Chen, Xiaokang, Nie, Xiaotao, Cheng, Xin, Liu, Xin, Xie, Xin, Liu, Xingchao, Yang, Xinyu, Li, Xinyuan, Su, Xuecheng, Lin, Xuheng, Li, X. Q., Jin, Xiangyue, Shen, Xiaojin, Chen, Xiaosha, Sun, Xiaowen, Wang, Xiaoxiang, Song, Xinnan, Zhou, Xinyi, Wang, Xianzu, Shan, Xinxia, Li, Y. K., Wang, Y. Q., Wei, Y. X., Zhang, Yang, Xu, Yanhong, Li, Yao, Zhao, Yao, Sun, Yaofeng, Wang, Yaohui, Yu, Yi, Zhang, Yichao, Shi, Yifan, Xiong, Yiliang, He, Ying, Piao, Yishi, Wang, Yisong, Tan, Yixuan, Ma, Yiyang, Liu, Yiyuan, Guo, Yongqiang, Ou, Yuan, Wang, Yuduan, Gong, Yue, Zou, Yuheng, He, Yujia, Xiong, Yunfan, Luo, Yuxiang, You, Yuxiang, Liu, Yuxuan, Zhou, Yuyang, Zhu, Y. X., Huang, Yanping, Li, Yaohui, Zheng, Yi, Zhu, Yuchen, Ma, Yunxian, Tang, Ying, Zha, Yukun, Yan, Yuting, Ren, Z. Z., Ren, Zehui, Sha, Zhangli, Fu, Zhe, Xu, Zhean, Xie, Zhenda, Zhang, Zhengyan, Hao, Zhewen, Ma, Zhicheng, Yan, Zhigang, Wu, Zhiyu, Gu, Zihui, Zhu, Zijia, Liu, Zijun, Li, Zilin, Xie, Ziwei, Song, Ziyang, Pan, Zizheng, Huang, Zhen, Xu, Zhipeng, Zhang, Zhongyu, and Zhang, Zhen
- Subjects
Computer Science - Computation and Language ,Computer Science - Artificial Intelligence ,Computer Science - Machine Learning - Abstract
We introduce our first-generation reasoning models, DeepSeek-R1-Zero and DeepSeek-R1. DeepSeek-R1-Zero, a model trained via large-scale reinforcement learning (RL) without supervised fine-tuning (SFT) as a preliminary step, demonstrates remarkable reasoning capabilities. Through RL, DeepSeek-R1-Zero naturally emerges with numerous powerful and intriguing reasoning behaviors. However, it encounters challenges such as poor readability, and language mixing. To address these issues and further enhance reasoning performance, we introduce DeepSeek-R1, which incorporates multi-stage training and cold-start data before RL. DeepSeek-R1 achieves performance comparable to OpenAI-o1-1217 on reasoning tasks. To support the research community, we open-source DeepSeek-R1-Zero, DeepSeek-R1, and six dense models (1.5B, 7B, 8B, 14B, 32B, 70B) distilled from DeepSeek-R1 based on Qwen and Llama.
- Published
- 2025
29. Pareto Set Learning for Multi-Objective Reinforcement Learning
- Author
-
Liu, Erlong, Wu, Yu-Chang, Huang, Xiaobin, Gao, Chengrui, Wang, Ren-Jian, Xue, Ke, and Qian, Chao
- Subjects
Computer Science - Machine Learning - Abstract
Multi-objective decision-making problems have emerged in numerous real-world scenarios, such as video games, navigation and robotics. Considering the clear advantages of Reinforcement Learning (RL) in optimizing decision-making processes, researchers have delved into the development of Multi-Objective RL (MORL) methods for solving multi-objective decision problems. However, previous methods either cannot obtain the entire Pareto front, or employ only a single policy network for all the preferences over multiple objectives, which may not produce personalized solutions for each preference. To address these limitations, we propose a novel decomposition-based framework for MORL, Pareto Set Learning for MORL (PSL-MORL), that harnesses the generation capability of hypernetwork to produce the parameters of the policy network for each decomposition weight, generating relatively distinct policies for various scalarized subproblems with high efficiency. PSL-MORL is a general framework, which is compatible for any RL algorithm. The theoretical result guarantees the superiority of the model capacity of PSL-MORL and the optimality of the obtained policy network. Through extensive experiments on diverse benchmarks, we demonstrate the effectiveness of PSL-MORL in achieving dense coverage of the Pareto front, significantly outperforming state-of-the-art MORL methods in the hypervolume and sparsity indicators., Comment: AAAI 2025 Accept
- Published
- 2025
30. A self-learning magnetic Hopfield neural network with intrinsic gradient descent adaption
- Author
-
Niu, Chang, Zhang, Huanyu, Xu, Chuanlong, Hu, Wenjie, Wu, Yunzhuo, Wu, Yu, Wang, Yadi, Wu, Tong, Zhu, Yi, Zhu, Yinyan, Wang, Wenbin, Wu, Yizheng, Yin, Lifeng, Xiao, Jiang, Yu, Weichao, Guo, Hangwen, and Shen, Jian
- Subjects
Condensed Matter - Disordered Systems and Neural Networks ,Physics - Applied Physics - Abstract
Physical neural networks using physical materials and devices to mimic synapses and neurons offer an energy-efficient way to implement artificial neural networks. Yet, training physical neural networks are difficult and heavily relies on external computing resources. An emerging concept to solve this issue is called physical self-learning that uses intrinsic physical parameters as trainable weights. Under external inputs (i.e. training data), training is achieved by the natural evolution of physical parameters that intrinsically adapt modern learning rules via autonomous physical process, eliminating the requirements on external computation resources.Here, we demonstrate a real spintronic system that mimics Hopfield neural networks (HNN) and unsupervised learning is intrinsically performed via the evolution of physical process. Using magnetic texture defined conductance matrix as trainable weights, we illustrate that under external voltage inputs, the conductance matrix naturally evolves and adapts Oja's learning algorithm in a gradient descent manner. The self-learning HNN is scalable and can achieve associative memories on patterns with high similarities. The fast spin dynamics and reconfigurability of magnetic textures offer an advantageous platform towards efficient autonomous training directly in materials., Comment: 21 pages, 5 figures
- Published
- 2025
- Full Text
- View/download PDF
31. Projected Spread Models
- Author
-
Ban, Jung-Chao, Hong, Jyy-I, Tsai, Cheng-Yu, and Wu, Yu-Liang
- Subjects
Mathematics - Dynamical Systems ,Mathematics - Probability - Abstract
We present a disease transmission model that considers both explicit and non-explicit factors. This approach is crucial for accurate prediction and control of infectious disease spread. In this paper, we extend the spread model from our previous works \cite{ban2021mathematical,ban2023randomspread, ban2023mathematical, ban2023spread} to a projected spread model that considers both hidden and explicit types. Additionally, we provide the spread rate for the projected spread model corresponding to the topological and random models. Furthermore, examples and numerical results are provided to illustrate the theory.
- Published
- 2025
32. Prophylactic Dexamethasone on Pain Flare in Spine Metastasis Stereotactic Body Radiotherapy and Hypofractionated Radiotherapy
- Author
-
WANG Runmei, YANG Hao, YU Rong, WANG Zhenfei, WU Yu, LI Hong, and HU Yue
- Subjects
dexamethasone ,spine metastasis ,stereotactic body radiation therapy ,pain flare ,Neoplasms. Tumors. Oncology. Including cancer and carcinogens ,RC254-282 - Abstract
Objective To explore the incidence of pain flare (PF) in spine metastasis stereotactic body radiotherapy (SBRT) or hypofractionated radiotherapy (HF) and the prophylactical effect of dexamethasone. Methods Sixty-five patients were treated with spine metastasis SBRT and randomly divided into control group (SBRT or HF, n=32) and treatment group (SBRT or HF and 4.5 mg dexamethasone, n=33). The brief pain inventory (BPI) was used to score the pain before, during and after treatment. PF was recorded and compared between two groups. Results The incidence of PF was 24.6% in all patients (control group: 37.5%, treatment group: 12.1%, P=0.018). PF in both group occurred in d1-2, accounting for 62.5% in all PF (control group: 66.7%, treatment group: 50%, P=0.551). The incidences of PF in control group were 66.7% and 33.3% for three and ten fractions scheme, respectively (P=0.001). However, the incidences of PF in treatment group were 50% and 50% for three and ten fractions scheme, respectively (P=0.643). Conclusion Oral dexamethasone has an excellent efficacy in prevention and treatment of PF in spine metastasis SBRT or HF, with significantly decreased incidence of PF. A phase Ⅲ clinical trial is required to finalize the optimal dose and schedule.
- Published
- 2021
- Full Text
- View/download PDF
33. Survey on Retrieval-based Chatbots
- Author
-
WU Yu, LI Zhou-jun
- Subjects
natural language processing ,chatbot ,text matching ,response selection ,pre-training technology ,Computer software ,QA76.75-76.765 ,Technology (General) ,T1-995 - Abstract
With the rapid progress of natural language processing techniques and the massive accessible conversational data on Internet,non-tasked oriented dialogue systems,also referred to as Chatbots,have achieved great success,and drawn attention from both academia and industry.Currently,there are two lines in chatbots research,retrieval-based chatbots and generation-based chatbots.Due to the fluent responses and low latency,retrieval-based chatbots is a common method in practice.This paperfirst briefly introduces the research background, basic structure and component modules of retrieval-based chatbots,and then illustrates the constraints of the response selection module and related data set in details.Subsequently,we summarize recent popular techniques for response selection problem,including:statistic method,representation-based neural network method,interaction-based neural network method,and pre-training-based method.Finally,we pose the challenges of chatbots and outline promising directions as future work.
- Published
- 2021
- Full Text
- View/download PDF
34. An Adaptive Mixer Allocation Algorithm for the Quantum Alternating Operator Ansatz
- Author
-
Ni, Xiao-Hui, Wu, Yu-Sen, Cai, Bin-Bin, Li, Wen-Min, Qin, Su-Juan, and Gao, Fei
- Subjects
Quantum Physics - Abstract
Recently, Hadfield et al. proposed the quantum alternating operator ansatz algorithm (QAOA+), an extension of the quantum approximate optimization algorithm (QAOA), to solve constrained combinatorial optimization problems (CCOPs). Compared with QAOA, QAOA+ enables the search for optimal solutions within a feasible solution space by encoding problem constraints into the mixer Hamiltonian, thus reducing the search space and eliminating the possibility of yielding infeasible solutions. However, QAOA+ may require substantial gates and circuit depth when the mixer is applied to all qubits in each layer, and the implementation cost of the mixer is high. To address these challenges, we propose the adaptive mixer allocation (AMA) algorithm. AMA constructs a feasible space by selectively applying the mixer to partial qubits in each layer based on the evolution of the optimization process. The performance of AMA in solving the maximum independent set (MIS) problem is investigated. Numerical simulation results demonstrate that, under the same number of optimization runs, AMA achieves a higher optimal approximation ratio--$1.82\%$ ($3.02\%$) higher than QAOA+ on ER (3-regular) graphs. Additionally, AMA significantly reduces the resource consumption, achieving only $34.08\%$ ($29.77\%$) of QAOA+ circuit depth and $15.05\%$ ($18.72\%$) of the number of CNOT gates, respectively, on ER (3-regular) graphs. These results highlight the ability of AMA to enhance computational efficiency and solution quality, making it particularly valuable for resource-constrained quantum devices.
- Published
- 2024
35. DeepSeek-V3 Technical Report
- Author
-
DeepSeek-AI, Liu, Aixin, Feng, Bei, Xue, Bing, Wang, Bingxuan, Wu, Bochao, Lu, Chengda, Zhao, Chenggang, Deng, Chengqi, Zhang, Chenyu, Ruan, Chong, Dai, Damai, Guo, Daya, Yang, Dejian, Chen, Deli, Ji, Dongjie, Li, Erhang, Lin, Fangyun, Dai, Fucong, Luo, Fuli, Hao, Guangbo, Chen, Guanting, Li, Guowei, Zhang, H., Bao, Han, Xu, Hanwei, Wang, Haocheng, Zhang, Haowei, Ding, Honghui, Xin, Huajian, Gao, Huazuo, Li, Hui, Qu, Hui, Cai, J. L., Liang, Jian, Guo, Jianzhong, Ni, Jiaqi, Li, Jiashi, Wang, Jiawei, Chen, Jin, Chen, Jingchang, Yuan, Jingyang, Qiu, Junjie, Li, Junlong, Song, Junxiao, Dong, Kai, Hu, Kai, Gao, Kaige, Guan, Kang, Huang, Kexin, Yu, Kuai, Wang, Lean, Zhang, Lecong, Xu, Lei, Xia, Leyi, Zhao, Liang, Wang, Litong, Zhang, Liyue, Li, Meng, Wang, Miaojun, Zhang, Mingchuan, Zhang, Minghua, Tang, Minghui, Li, Mingming, Tian, Ning, Huang, Panpan, Wang, Peiyi, Zhang, Peng, Wang, Qiancheng, Zhu, Qihao, Chen, Qinyu, Du, Qiushi, Chen, R. J., Jin, R. L., Ge, Ruiqi, Zhang, Ruisong, Pan, Ruizhe, Wang, Runji, Xu, Runxin, Zhang, Ruoyu, Chen, Ruyi, Li, S. S., Lu, Shanghao, Zhou, Shangyan, Chen, Shanhuang, Wu, Shaoqing, Ye, Shengfeng, Ma, Shirong, Wang, Shiyu, Zhou, Shuang, Yu, Shuiping, Zhou, Shunfeng, Pan, Shuting, Wang, T., Yun, Tao, Pei, Tian, Sun, Tianyu, Xiao, W. L., Zeng, Wangding, Zhao, Wanjia, An, Wei, Liu, Wen, Liang, Wenfeng, Gao, Wenjun, Yu, Wenqin, Zhang, Wentao, Li, X. Q., Jin, Xiangyue, Wang, Xianzu, Bi, Xiao, Liu, Xiaodong, Wang, Xiaohan, Shen, Xiaojin, Chen, Xiaokang, Zhang, Xiaokang, Chen, Xiaosha, Nie, Xiaotao, Sun, Xiaowen, Wang, Xiaoxiang, Cheng, Xin, Liu, Xin, Xie, Xin, Liu, Xingchao, Yu, Xingkai, Song, Xinnan, Shan, Xinxia, Zhou, Xinyi, Yang, Xinyu, Li, Xinyuan, Su, Xuecheng, Lin, Xuheng, Li, Y. K., Wang, Y. Q., Wei, Y. X., Zhu, Y. X., Zhang, Yang, Xu, Yanhong, Huang, Yanping, Li, Yao, Zhao, Yao, Sun, Yaofeng, Li, Yaohui, Wang, Yaohui, Yu, Yi, Zheng, Yi, Zhang, Yichao, Shi, Yifan, Xiong, Yiliang, He, Ying, Tang, Ying, Piao, Yishi, Wang, Yisong, Tan, Yixuan, Ma, Yiyang, Liu, Yiyuan, Guo, Yongqiang, Wu, Yu, Ou, Yuan, Zhu, Yuchen, Wang, Yuduan, Gong, Yue, Zou, Yuheng, He, Yujia, Zha, Yukun, Xiong, Yunfan, Ma, Yunxian, Yan, Yuting, Luo, Yuxiang, You, Yuxiang, Liu, Yuxuan, Zhou, Yuyang, Wu, Z. F., Ren, Z. Z., Ren, Zehui, Sha, Zhangli, Fu, Zhe, Xu, Zhean, Huang, Zhen, Zhang, Zhen, Xie, Zhenda, Zhang, Zhengyan, Hao, Zhewen, Gou, Zhibin, Ma, Zhicheng, Yan, Zhigang, Shao, Zhihong, Xu, Zhipeng, Wu, Zhiyu, Zhang, Zhongyu, Li, Zhuoshu, Gu, Zihui, Zhu, Zijia, Liu, Zijun, Li, Zilin, Xie, Ziwei, Song, Ziyang, Gao, Ziyi, and Pan, Zizheng
- Subjects
Computer Science - Computation and Language ,Computer Science - Artificial Intelligence - Abstract
We present DeepSeek-V3, a strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token. To achieve efficient inference and cost-effective training, DeepSeek-V3 adopts Multi-head Latent Attention (MLA) and DeepSeekMoE architectures, which were thoroughly validated in DeepSeek-V2. Furthermore, DeepSeek-V3 pioneers an auxiliary-loss-free strategy for load balancing and sets a multi-token prediction training objective for stronger performance. We pre-train DeepSeek-V3 on 14.8 trillion diverse and high-quality tokens, followed by Supervised Fine-Tuning and Reinforcement Learning stages to fully harness its capabilities. Comprehensive evaluations reveal that DeepSeek-V3 outperforms other open-source models and achieves performance comparable to leading closed-source models. Despite its excellent performance, DeepSeek-V3 requires only 2.788M H800 GPU hours for its full training. In addition, its training process is remarkably stable. Throughout the entire training process, we did not experience any irrecoverable loss spikes or perform any rollbacks. The model checkpoints are available at https://github.com/deepseek-ai/DeepSeek-V3.
- Published
- 2024
36. On character values of $GL_n(\mathbb F_q)$
- Author
-
Jing, Naihuan and Wu, Yu
- Subjects
Mathematics - Representation Theory ,Mathematics - Group Theory ,Mathematics - Quantum Algebra ,Primary: 20C33, 17B69, Secondary: 05E10 - Abstract
In this paper, we use vertex operator techniques to compute character values on unipotent classes of $\GL_n(\mathbb F_q)$. By realizing the Grothendieck ring $R_G=\bigoplus_{n\geq0}^\infty R(\GL_n(\mathbb F_q))$ as Fock spaces, we formulate the Murnanghan-Nakayama rule of $\GL_n(\mathbb F_q)$ between Schur functions colored by an orbit $\phi$ of linear characters of $\overline{\mathbb F}_q$ under the Frobenius automorphism on and modified Hall-Littlewood functions colored by $f_1=t-1$, which provides detailed information on the character table of $\GL_n(\mathbb F_q)$. As applications, we use vertex algebraic methods to determine the Steinberg characters of $\GL_n(\mathbb F_q)$, which were previously determined by Curtis-Lehrer-Tits via geometry of homology groups of spherical buildings and Springer-Zelevinsky utilizing Hopf algebras., Comment: 15pp
- Published
- 2024
37. DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding
- Author
-
Wu, Zhiyu, Chen, Xiaokang, Pan, Zizheng, Liu, Xingchao, Liu, Wen, Dai, Damai, Gao, Huazuo, Ma, Yiyang, Wu, Chengyue, Wang, Bingxuan, Xie, Zhenda, Wu, Yu, Hu, Kai, Wang, Jiawei, Sun, Yaofeng, Li, Yukun, Piao, Yishi, Guan, Kang, Liu, Aixin, Xie, Xin, You, Yuxiang, Dong, Kai, Yu, Xingkai, Zhang, Haowei, Zhao, Liang, Wang, Yisong, and Ruan, Chong
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Computer Science - Artificial Intelligence ,Computer Science - Computation and Language - Abstract
We present DeepSeek-VL2, an advanced series of large Mixture-of-Experts (MoE) Vision-Language Models that significantly improves upon its predecessor, DeepSeek-VL, through two key major upgrades. For the vision component, we incorporate a dynamic tiling vision encoding strategy designed for processing high-resolution images with different aspect ratios. For the language component, we leverage DeepSeekMoE models with the Multi-head Latent Attention mechanism, which compresses Key-Value cache into latent vectors, to enable efficient inference and high throughput. Trained on an improved vision-language dataset, DeepSeek-VL2 demonstrates superior capabilities across various tasks, including but not limited to visual question answering, optical character recognition, document/table/chart understanding, and visual grounding. Our model series is composed of three variants: DeepSeek-VL2-Tiny, DeepSeek-VL2-Small and DeepSeek-VL2, with 1.0B, 2.8B and 4.5B activated parameters respectively. DeepSeek-VL2 achieves competitive or state-of-the-art performance with similar or fewer activated parameters compared to existing open-source dense and MoE-based models. Codes and pre-trained models are publicly accessible at https://github.com/deepseek-ai/DeepSeek-VL2.
- Published
- 2024
38. The Silent Prompt: Initial Noise as Implicit Guidance for Goal-Driven Image Generation
- Author
-
Wang, Ruoyu, Huang, Huayang, Zhu, Ye, Russakovsky, Olga, and Wu, Yu
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Text-to-image synthesis (T2I) has advanced remarkably with the emergence of large-scale diffusion models. In the conventional setup, the text prompt provides explicit, user-defined guidance, directing the generation process by denoising a randomly sampled Gaussian noise. In this work, we reveal that the often-overlooked noise itself encodes inherent generative tendencies, acting as a "silent prompt" that implicitly guides the output. This implicit guidance, embedded in the noise scheduler design of diffusion model formulations and their training stages, generalizes across a wide range of T2I models and backbones. Building on this insight, we introduce NoiseQuery, a novel strategy that selects optimal initial noise from a pre-built noise library to meet diverse user needs. Our approach not only enhances high-level semantic alignment with text prompts, but also allows for nuanced adjustments of low-level visual attributes, such as texture, sharpness, shape, and color, which are typically challenging to control through text alone. Extensive experiments across various models and target attributes demonstrate the strong performance and zero-shot transferability of our approach, requiring no additional optimization., Comment: 18 pages, 18 figures, 6 tables
- Published
- 2024
39. Neural Network-Based Frequency Optimization for Superconducting Quantum Chips
- Author
-
Lu, Bin-Han, Wang, Peng, Li, Qing-Song, Wu, Yu-Chun, Chen, Zhao-Yun, and Guo, Guo-Ping
- Subjects
Quantum Physics - Abstract
Optimizing the frequency configuration of qubits and quantum gates in superconducting quantum chips presents a complex NP-complete optimization challenge. This process is critical for enabling practical control while minimizing decoherence and suppressing significant crosstalk. In this paper, we propose a neural network-based frequency configuration approach. A trained neural network model estimates frequency configuration errors, and an intermediate optimization strategy identifies optimal configurations within localized regions of the chip. The effectiveness of our method is validated through randomized benchmarking and cross-entropy benchmarking. Furthermore, we design a crosstalk-aware hardware-efficient ansatz for variational quantum eigensolvers, achieving improved energy computations.
- Published
- 2024
40. Determination of 4 categories of flavor substances in liquor by gas chromatography mass spectrometry
- Author
-
XU Zhi-fei, WU Yu-kang, JIANG Yu-hong, and ZHOU Run
- Subjects
gas chromatography-mass spectrometry ,liquor ,flavoring substances ,flavor type ,Food processing and manufacture ,TP368-456 - Abstract
Objective: To establish an analytical method for the determination of four main flavor substances in liquor. Methods: A gas chromatography mass spectrometry (GC-MS) method was established for the simultaneous determination of esters, alcohols, acids, and aldehydes in liquor. Single ion monitoring (SIM) mode was chosen. Then the method was inspected through comparison with the standard mass spectrum database, quantitative analysis of the external standard method and relative standard deviation of day-in and day-out were investigated. Results: The limits of detection of the four types of flavor compounds in liquor were 0.007~0.070 mg/L, and the limits of quantitation were 0.020~0.200 mg/L, and the linearities were satisfactory in the range of 0.000 5~5.0 g/L, with the correlation coefficients of R2≥0.998 5. The accuracy was tested by adding three levels of low, medium and high concentrations of standard substances to ethanol in water (60%), and the average recoveries of the four categories of flavor substances were 87.8%~107.4%, with the relative standard deviations (RSDs) of 2.0%~7.4% (n=6). Conclusion: The method is accurate and reliable, with good repeatability and high sensitivity. The method is stable and can be used for the determination of various trace substances in wine. It was found that the volatile flavor compounds of different flavor liquors were different to some extent, which could provide a reference for further research on the flavor compounds of liquor.
- Published
- 2022
- Full Text
- View/download PDF
41. Nanostructures of self-assembled and co-assembled chiral diphenylalanine
- Author
-
SHI Jiali, HOU Yangqian, ZHU Zhongjie, WU Yu, ZHAO Hongwei, and ZHANG Yi
- Subjects
diphenylalanine ,self-assembly ,co-assembly ,xrd ,nanostructures ,Nuclear engineering. Atomic power ,TK9001-9401 - Abstract
BackgroundAs a fundamental property of peptide molecules, chirality has been demonstrated to play an important role in controlling the structures and characteristics of peptide supramolecular systems. However, the mechanism through which chirality takes effect has not been clarified.PurposeThis study aims to examine the self- and co-assembled nanostructures and analyze the intermolecular interactions that drive the assembly by employing diphenylalanine (FF), along with the core recognition sequence of Amyloid-β protein (Aβ) and its enantiomer D-Phe-D-Phe (ff), in a model system.MethodsA series of structural and morphological analyses were conducted in the experiments. First, the scanning electron microscopy (SEM) and atomic force microscopy (AFM) images of the assembled nanostructures were obtained to observe the microscopic morphology and topological structure of the assembled FF and ff. Subsequently, circular dichroism (CD) and Fourier transform infrared spectroscopy (FTIR) were employed to characterize the secondary structures of peptides in the nanostructures. Finally, fluorescence emission spectrum and X-ray diffraction (XRD) analyses revealed the intermolecular interactions between peptide and solvent molecules.ResultsThe findings demonstrate that FF and ff self-assemble into similar fibrous nanostructures, and their chirality primarily affects the interactions between peptide molecules, as well as those between peptide and water molecules. Furthermore, the formation of new crystalline phases for the co-assembly of FF and ff was confirmed by XRD.ConclusionsOur results may facilitate the understanding of the formation mechanism of amyloid fibers and design of peptide supramolecular materials.
- Published
- 2023
- Full Text
- View/download PDF
42. Improving flexural strength of UHPC with sustainably synthesized graphene oxide
- Author
-
Luo Qizhi, Wu Yu-You, Qiu Wenjun, Huang Haoliang, Pei Songfeng, Lambert Paul, and Hui David
- Subjects
sustainably synthesized graphene oxide ,ultrahigh-performance concrete ,flexural strength ,Technology ,Chemical technology ,TP1-1185 ,Physical and theoretical chemistry ,QD450-801 - Published
- 2021
- Full Text
- View/download PDF
43. Fabrication and characterization of cellulose/ lignin microspheres films with UV-blocking
- Author
-
WAN Yu-ling, HU Yu-lu, XU Du-xin, HUANG Jian-bo, XU Feng, WU Yu-ying, and ZHANG Xue-ming
- Subjects
lignin microsphere ,cellulose film ,anti-ultraviolet function ,packaging material ,Materials of engineering and construction. Mechanics of materials ,TA401-492 - Abstract
With the rapid development of packaging industry and the improvement of environmental protection requirements of human society, biodegradable functional packaging film materials have attracted great attention. However, commercial biodegradable functional package film materials have been constrained by its high cost, low mechanical properties and water resistance. The lignin microspheres were firstly prepared by self-assembling and were adhered to the surface of cellulose film via self-depositing, resulting in fabrication of novel cellulose-based UV-blocking film materials. Subsequently, the surface morphology, ultraviolet resistance and mechanical properties of the obtained functional film materials were studied. The surface properties of films were investigated by scanning microscopy (SEM),infrared spectroscopy (FTIR) and confocal laser scanning microscopy.The mechanical properties and UV resistance of cellulose based films were characterized by tensile test and UV transmittance test.It is noteworthy that the lignin microspheres with 1-2 μm in size are uniformly distributed on cellulose films and hydrophobic modification of cellulose-based films facilitated the deposition of lignin. Moreover, the deposited contents of lignin microspheres increase with the increase of the lignin concentration in solution. Due to the introduction of lignin microspheres, the strength of cellulose composites film increases by 22% and the UVB-shielding capacity reaches 94% as compared with the controlled cellulose films.
- Published
- 2021
- Full Text
- View/download PDF
44. Advance on the dispersion treatment of graphene oxide and the graphene oxide modified cement-based materials
- Author
-
Liu Changjiang, Huang Xiaochuan, Wu Yu-You, Deng Xiaowei, Zheng Zhoulian, Xu Zhong, and Hui David
- Subjects
dispersion treatment ,graphene oxide ,cement-based materials ,development prospects ,nanomaterials ,Technology ,Chemical technology ,TP1-1185 ,Physical and theoretical chemistry ,QD450-801 - Abstract
For the high demand for cement-based materials in buildings, improving the performance of cement-based materials has become the focus of relevant researchers. In recent years, nanomaterials have broad prospects in many fields such as architecture by virtue of their “lightweight, high strength, and strong solidity” characteristics. As a modifier of cement-based materials, it has also become a research hotspot. Graphene oxide (GO) is one of the most representative graphene-based nanomaterials. Because of its extremely high specific surface area and excellent physical properties, it has greatly improved the properties of cement-based materials. GO acts as an enhancer of cement composites that brings people unlimited imagination. The research progress of GO-modified cement-based materials is reviewed. The purpose is to point out the limitations of current research and provide a reference for later research. The dispersion treatment of GO and the properties of its modified cement-based materials are analyzed and summarized. In addition, the further research work that is needed and future development prospect are discussed.
- Published
- 2021
- Full Text
- View/download PDF
45. SOWing Information: Cultivating Contextual Coherence with MLLMs in Image Generation
- Author
-
Pei, Yuhan, Wang, Ruoyu, Yang, Yongqi, Zhu, Ye, Russakovsky, Olga, and Wu, Yu
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Computer Science - Artificial Intelligence - Abstract
Originating from the diffusion phenomenon in physics, which describes the random movement and collisions of particles, diffusion generative models simulate a random walk in the data space along the denoising trajectory. This allows information to diffuse across regions, yielding harmonious outcomes. However, the chaotic and disordered nature of information diffusion in diffusion models often results in undesired interference between image regions, causing degraded detail preservation and contextual inconsistency. In this work, we address these challenges by reframing disordered diffusion as a powerful tool for text-vision-to-image generation (TV2I) tasks, achieving pixel-level condition fidelity while maintaining visual and semantic coherence throughout the image. We first introduce Cyclic One-Way Diffusion (COW), which provides an efficient unidirectional diffusion framework for precise information transfer while minimizing disruptive interference. Building on COW, we further propose Selective One-Way Diffusion (SOW), which utilizes Multimodal Large Language Models (MLLMs) to clarify the semantic and spatial relationships within the image. Based on these insights, SOW combines attention mechanisms to dynamically regulate the direction and intensity of diffusion according to contextual relationships. Extensive experiments demonstrate the untapped potential of controlled information diffusion, offering a path to more adaptive and versatile generative models in a learning-free manner., Comment: Project page: https://pyh-129.github.io/SOW/
- Published
- 2024
46. Rethinking the Intermediate Features in Adversarial Attacks: Misleading Robotic Models via Adversarial Distillation
- Author
-
Zhao, Ke, Huang, Huayang, Li, Miao, and Wu, Yu
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence ,Computer Science - Robotics - Abstract
Language-conditioned robotic learning has significantly enhanced robot adaptability by enabling a single model to execute diverse tasks in response to verbal commands. Despite these advancements, security vulnerabilities within this domain remain largely unexplored. This paper addresses this gap by proposing a novel adversarial prompt attack tailored to language-conditioned robotic models. Our approach involves crafting a universal adversarial prefix that induces the model to perform unintended actions when added to any original prompt. We demonstrate that existing adversarial techniques exhibit limited effectiveness when directly transferred to the robotic domain due to the inherent robustness of discretized robotic action spaces. To overcome this challenge, we propose to optimize adversarial prefixes based on continuous action representations, circumventing the discretization process. Additionally, we identify the beneficial impact of intermediate features on adversarial attacks and leverage the negative gradient of intermediate self-attention features to further enhance attack efficacy. Extensive experiments on VIMA models across 13 robot manipulation tasks validate the superiority of our method over existing approaches and demonstrate its transferability across different model variants.
- Published
- 2024
47. Weak Host Interactions Induced Thermal Transport Properties of Metal Halide Perovskites Deviating from the Rattling Model
- Author
-
Wu, Yu, Ji, Linxuan, Zeng, Shuming, Ding, Yimin, and Zhou, Liujiang
- Subjects
Condensed Matter - Materials Science - Abstract
The low-frequency phonon branches of metal halide perovskites typically exhibit the characteristic of hardening with the increase of the cation mass, which leads to anomalous thermal transport phenomenon. However, the underlying physical mechanism is not yet understood. Here, we theoretically compare the thermal transport properties of $A_2$SnI$_6$ ($A$=K, Rb, and Cs) perovskites. The thermal transport in perovskites is widely explained using the rattling model, where ``guest'' cations inside the metal halide framework act as ``rattlers'', but this fails to explain the following phenomenon: The low-frequency phonon branch of $A_2$SnI$_6$ perovskites is insensitive to the mass of the $A^+$ cation and strongly correlated with the interaction of the $A^+$ cation with the I$^-$ anion in the octahedral structures. The failure of the rattling model stems mainly from the weak interactions between the octahedral structures. By developing a new spring model, we successfully explain the thermal transport behavior in $A_2$SnI$_6$ perovskites. Our work gives new insights into the thermal transport mechanism in metal halide perovskites, which has a guiding significance for designing extremely low thermal conductivity materials.
- Published
- 2024
48. EAPCR: A Universal Feature Extractor for Scientific Data without Explicit Feature Relation Patterns
- Author
-
Yu, Zhuohang, An, Ling, Li, Yansong, Wu, Yu, Dong, Zeyu, Liu, Zhangdi, Gao, Le, Zhang, Zhenyu, and Zhou, Chichun
- Subjects
Computer Science - Machine Learning ,Computer Science - Computer Vision and Pattern Recognition - Abstract
Conventional methods, including Decision Tree (DT)-based methods, have been effective in scientific tasks, such as non-image medical diagnostics, system anomaly detection, and inorganic catalysis efficiency prediction. However, most deep-learning techniques have struggled to surpass or even match this level of success as traditional machine-learning methods. The primary reason is that these applications involve multi-source, heterogeneous data where features lack explicit relationships. This contrasts with image data, where pixels exhibit spatial relationships; textual data, where words have sequential dependencies; and graph data, where nodes are connected through established associations. The absence of explicit Feature Relation Patterns (FRPs) presents a significant challenge for deep learning techniques in scientific applications that are not image, text, and graph-based. In this paper, we introduce EAPCR, a universal feature extractor designed for data without explicit FRPs. Tested across various scientific tasks, EAPCR consistently outperforms traditional methods and bridges the gap where deep learning models fall short. To further demonstrate its robustness, we synthesize a dataset without explicit FRPs. While Kolmogorov-Arnold Network (KAN) and feature extractors like Convolutional Neural Networks (CNNs), Graph Convolutional Networks (GCNs), and Transformers struggle, EAPCR excels, demonstrating its robustness and superior performance in scientific tasks without FRPs.
- Published
- 2024
49. Quantum Homotopy Analysis Method with Secondary Linearization for Nonlinear Partial Differential Equations
- Author
-
Xue, Cheng, Xu, Xiao-Fan, Zhuang, Xi-Ning, Sun, Tai-Ping, Wang, Yun-Jie, Tan, Ming-Yang, Ye, Chuang-Chao, Liu, Huan-Yu, Wu, Yu-Chun, Chen, Zhao-Yun, and Guo, Guo-Ping
- Subjects
Quantum Physics - Abstract
Nonlinear partial differential equations (PDEs) are crucial for modeling complex fluid dynamics and are foundational to many computational fluid dynamics (CFD) applications. However, solving these nonlinear PDEs is challenging due to the vast computational resources they demand, highlighting the pressing need for more efficient computational methods. Quantum computing offers a promising but technically challenging approach to solving nonlinear PDEs. Recently, Liao proposed a framework that leverages quantum computing to accelerate the solution of nonlinear PDEs based on the homotopy analysis method (HAM), a semi-analytical technique that transforms nonlinear PDEs into a series of linear PDEs. However, the no-cloning theorem in quantum computing poses a major limitation, where directly applying quantum simulation to each HAM step results in exponential complexity growth with the HAM truncation order. This study introduces a "secondary linearization" approach that maps the whole HAM process into a system of linear PDEs, allowing for a one-time solution using established quantum PDE solvers. Our method preserves the exponential speedup of quantum linear PDE solvers while ensuring that computational complexity increases only polynomially with the HAM truncation order. We demonstrate the efficacy of our approach by applying it to the Burgers' equation and the Korteweg-de Vries (KdV) equation. Our approach provides a novel pathway for transforming nonlinear PDEs into linear PDEs, with potential applications to fluid dynamics. This work thus lays the foundation for developing quantum algorithms capable of solving the Navier-Stokes equations, ultimately offering a promising route to accelerate their solutions using quantum computing., Comment: 22 pages, 4 figures
- Published
- 2024
50. OCMDP: Observation-Constrained Markov Decision Process
- Author
-
Wang, Taiyi, Liu, Jianheng, Lee, Bryan, Wu, Zhihao, and Wu, Yu
- Subjects
Computer Science - Artificial Intelligence ,Computer Science - Machine Learning ,Electrical Engineering and Systems Science - Systems and Control - Abstract
In many practical applications, decision-making processes must balance the costs of acquiring information with the benefits it provides. Traditional control systems often assume full observability, an unrealistic assumption when observations are expensive. We tackle the challenge of simultaneously learning observation and control strategies in such cost-sensitive environments by introducing the Observation-Constrained Markov Decision Process (OCMDP), where the policy influences the observability of the true state. To manage the complexity arising from the combined observation and control actions, we develop an iterative, model-free deep reinforcement learning algorithm that separates the sensing and control components of the policy. This decomposition enables efficient learning in the expanded action space by focusing on when and what to observe, as well as determining optimal control actions, without requiring knowledge of the environment's dynamics. We validate our approach on a simulated diagnostic task and a realistic healthcare environment using HeartPole. Given both scenarios, the experimental results demonstrate that our model achieves a substantial reduction in observation costs on average, significantly outperforming baseline methods by a notable margin in efficiency., Comment: Full paper, 14 Pages
- Published
- 2024
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.