A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Filters
Markov Boundary Learning with Streaming Data for Supervised Classification
2020
IEEE Access
Preserved Fulltext
KUI YU received the Ph.D. degree in computer science from the Hefei University of Technology, Hefei, China, in 2013. ...
doi:10.1109/access.2020.2998482
fatcat:3gnbu5i37fextiolallc3tndeu
Secure Collaborative Training and Inference for XGBoost
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
In recent years, gradient boosted decision tree learning has proven to be an effective method of training robust models. Moreover, collaborative learning among multiple parties has the potential to greatly benefit all parties involved, but organizations have also encountered obstacles in sharing sensitive data due to business, regulatory, and liability concerns. We propose Secure XGBoost, a privacy-preserving system that enables multiparty training and inference of XGBoost models. Secure
arXiv:2010.02524v1
fatcat:mh5gmhwqefe7lfh25jet2jgpzq
more »
... protects the privacy of each party's data as well as the integrity of the computation with the help of hardware enclaves. Crucially, Secure XGBoost augments the security of the enclaves using novel data-oblivious algorithms that prevent access side-channel attacks on enclaves induced via access pattern leakage.
S3ML: A Secure Serving System for Machine Learning Inference
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
We present S3ML, a secure serving system for machine learning inference in this paper. S3ML runs machine learning models in Intel SGX enclaves to protect users' privacy. S3ML designs a secure key management service to construct flexible privacy-preserving server clusters and proposes novel SGX-aware load balancing and scaling methods to satisfy users' Service-Level Objectives. We have implemented S3ML based on Kubernetes as a low-overhead, high-available, and scalable system. We demonstrate the
arXiv:2010.06212v1
fatcat:bb5krzimovfjtbdmvhxkb2snoa
more »
... system performance and effectiveness of S3ML through extensive experiments on a series of widely-used models.
Dynamic thermal coupling modeling and analysis of wet electro-hydrostatic actuator
2021
Chinese Journal of Aeronautics
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
The electro-hydrostatic actuator (EHA) used in more electric aircraft (MEA) has been extensively studied due to its advantages of high reliability and high integration. However, this high integration results in a small heat dissipation area, leading to high-temperature problems. Generally, to reduce the temperature, a wet cooling method of using the pump leakage oil to cool the motor is adopted, which can also increase the difficulty of accurately predicting the system temperature in the early
doi:10.1016/j.cja.2021.04.007
fatcat:fcnsqadb5vhajcmnr6aavuqinu
more »
... esign stage. To solve this problem, a dynamic coupling thermal model of a wet EHA is proposed in this paper. In particular, the leakage oil of the pump is used as a coupling item between the electrical system and the hydraulic system. Then, an improved T-equivalent block model is proposed to address the uneven distribution of axial oil temperature inside the motor, and the control node method is applied to hydraulic system thermal modeling. Meanwhile, a dynamic coupling thermal model is developed that enables a dynamic evaluation of the wet EHA temperature. Then, experimental prototypes of wet motor and wet EHA are developed, while the temperature response of the wet motor at different rotation speeds and different loads and the temperature response of the wet EHA at no-load condition were verified experimentally at room temperature, respectively. The maximum temperature difference between the experimental and theoretical results of the wet motor as well as the experimental and theoretical results of the wet EHA is less than 8°C. These test results indicate that the dynamic coupling thermal model is valid and demonstrate that the thermal coupling modeling method proposed in this paper can provide a basis for the detailed thermal design of EHA.
A Hybrid-Domain Framework for Secure Gradient Tree Boosting
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Other Versions
Gradient tree boosting (e.g. XGB) is one of the most widely usedmachine learning models in practice. How to build a secure XGB inface of data isolation problem becomes a hot research topic. However, existing works tend to leak intermediate information and thusraise potential privacy risk. In this paper, we propose a novel framework for two parties to build secure XGB with vertically partitioneddata. Specifically, we associate Homomorphic Encryption (HE) domain with Secret Sharing (SS) domain by
arXiv:2005.08479v1
fatcat:2vbshbawhjhmxgyj4lhwooepdq
more »
... providing the two-waytransformation primitives. The framework generally promotes theefficiency for privacy preserving machine learning and offers theflexibility to implement other machine learning models. Then weelaborate two secure XGB training algorithms as well as a corresponding prediction algorithm under the hybrid security domains.Next, we compare our proposed two training algorithms throughboth complexity analysis and experiments. Finally, we verify themodel performance on benchmark dataset and further apply ourwork to a real-world scenario.
Near-Infrared Imaging for Information Embedding and Extraction with Layered Structures
2022
ACM Transactions on Graphics
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
ACKNOWLEDGMENTS Weiwei Jiang, Difeng Yu and Chaofan Wang are supported by Melbourne Research Scholarships. Zhanna Sarsenbayeva is supported by a Doreen Thomas Postdoctoral Fellowship. ...
doi:10.1145/3533426
fatcat:5xjwtuu5zbc2rbwtbfaalcxjhq
Trajectory Generation and Optimization Using the Mutual Learning and Adaptive Ant Colony Algorithm in Uneven Environments
2022
Applied Sciences
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Aiming at the trajectory generation and optimization of mobile robots in complex and uneven environments, a hybrid scheme using mutual learning and adaptive ant colony optimization (MuL-ACO) is proposed in this paper. In order to describe the uneven environment with various obstacles, a 2D-H map is introduced in this paper. Then an adaptive ant colony algorithm based on simulated annealing (SA) is proposed to generate initial trajectories of mobile robots, where based on a de-temperature
doi:10.3390/app12094629
fatcat:db2xuppmcfabfej6i7x7txkupm
more »
... n of the simulated annealing algorithm, the pheromone volatilization factor is adaptively adjusted to accelerate the convergence of the algorithm. Moreover, the length factor, height factor, and smooth factor are considered in the comprehensive heuristic function of ACO to adapt to uneven environments. Finally, a mutual learning algorithm is designed to further smooth and shorten initial trajectories, in which different trajectory node sequences learn from each other to acquire the shortest trajectory sequence to optimize the trajectory. In order to verify the effectiveness of the proposed scheme, MuL-ACO is compared with several well-known and novel algorithms in terms of running time, trajectory length, height, and smoothness. The experimental results show that MuL-ACO can generate a collision-free trajectory with a high comprehensive quality in uneven environments.
Prognostic Value of Elevated Cardiac Troponin I After Aneurysmal Subarachnoid Hemorrhage
2021
Frontiers in Neurology
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Object: Patients with aneurysmal subarachnoid hemorrhage (aSAH) have an increased incidence of cardiac events and short-term unfavorable neurological outcomes during the acute phase of bleeding. We studied whether troponin I elevation after ictus can predict future major adverse cardiac events (MACEs) and long-term neurological outcomes after 2 years.Methods: Consecutive aSAH patients within 3 days of bleeding were eligible for review from a prospective observational cohort (ClinicalTrials.gov
doi:10.3389/fneur.2021.677961
pmid:34135855
pmcid:PMC8200557
fatcat:57jeudq6jveoxmewmboeqzcqri
more »
... dentifier: NCT04785976). Potential predictors of future MACEs and unfavorable long-term neurological outcomes were calculated by Cox and logistic regression analyses. Additional Kaplan–Meier curves were performed.Results: A total of 213 patients were enrolled with an average follow-up duration of 34.3 months. Individuals were divided into two groups: elevated cTnI group and unelevated cTnI group. By the last available follow-up, 20 patients had died, with an overall all-cause mortality rate of 9.4% and an annual all-cause mortality rate of 3.8%. Patients with elevated cTnI had a significantly higher risk of future MACEs (10.6 vs. 2.1%, p = 0.024, and 95% CI: 1.256–23.875) and unfavorable neurological outcomes at discharge, 3-month, 1-, 2-years, and last follow-up (p = 0.001, p < 0.001, p = 0.001, p < 0.001, and p < 0.001, respectively). In the Cox analysis for future MACE, elevated cTnI was the only independent predictor (HR = 5.980; 95% CI: 1.428–25.407, and p = 0.014). In the multivariable logistic analysis for unfavorable neurological outcomes, peak cTnI was significant (OR = 2.951; 95% CI: 1.376–6.323; p = 0.005). Kaplan–Meier analysis indicated that the elevated cTnI was correlated with future MACE (log-rank test, p = 0.007) and subsequent death (log-rank test, p = 0.004).Conclusion: cTnI elevation after aSAH could predict future MACEs and unfavorable neurological outcomes.
High Temperature Flexural Deformation Properties of Engineered Cementitious Composites (ECC) with Hybrid Fiber Reinforcement
2020
Research and Application of Materials Science
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Author Contributions: Zhihui YU performed the experiment and wrote this manuscript; Zhen YUAN contributed significantly to manuscript preparation; Chaofan XIA performed the data analyses; Cong ZHANG helped ...
doi:10.33142/rams.v2i2.3168
fatcat:4msbpvzqc5gcbjekvhspx6aw3m
Taklimakan desert carbon-sink decreases under climate change
2019
Science Bulletin
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
The global carbon balance is a core issue in climate change research and a focus of international policy concern [1-3]. The "missing carbon sink" caused by approximately 1.6-2.0 Pg C a À1 (1 Pg = 10 15 g) that is currently unaccounted for, has long plagued researchers [4, 5] . Evidence is mounting those seemingly lifeless desert ecosystems, whose roles in the global carbon-cycle have long been neglected, exhibit the unconventional phenomenon of absorption of atmospheric CO 2 , sequestering
doi:10.1016/j.scib.2019.12.022
fatcat:x67dztlxufdwfp3nudxkp2bldq
more »
... ous amounts of CO 2 and thereby creating a significant carbon-sink [6] [7] [8] [9] . This scenario appears to help narrow the gap in the missing carbonsink. However, this view has been questioned for the inaccurate carbon sequestration location [10] . The effects of abiotic processes, such as soil temperature gradient, moisture content, parent material, salt/alkali, and pH, on desert CO 2 exchange, have gradually revealed that the phenomenon of a desert carbon-sink is intimately associated with abiotic processes [11] . Thus far, the main processes, whereby, deserts act as carbon-sinks may include the following: (1) variation in the volume of gases caused by changes in pressure and temperature; (2) changes in solubility of CO 2 in soil water films; (3) pH-mediated CO 2 dissolution chemistry; and (4) surface adhesion of CO 2 onto soil minerals [11] [12] [13] . In ecosystems with relatively high productivity, the effects of abiotic processes on CO 2 exchange are often neglected because of strong biological respiration. In desert ecosystems, however, the effects of abiotic processes on CO 2 exchange cannot be ignored, although the CO 2 flux is extremely weak. The effects of abiotic processes are even greater than the effects of biological processes [11, 14] . The latest research indicates that the absorbed CO 2 gradually, enters the desert groundwater layer by the process of leaching, and eventually converge in the underground saline water layer under the vast desert with the groundwater movement. This process is similar to the inorganic carbon sinks in the ocean [9] . The Taklimakan Desert (TD) is the world's second-largest shifting desert and plays a vital role in the carbon-sink properties of desert ecosystems. Benefit by the abundant saline-alkali groundwater and extensive salinization, the TD has a huge carbon storage capacity. However, there is uncertainty regarding the magnitude of the contribution made by each internal process towards the TD carbon-sink; and how these processes will respond to climate change. The TD land surface mainly consists of continuous shifting sand cover, which shows the high homogeneity of this desert. Therefore, the hinterland of the TD can be considered the most representative research area in this desert. In the hinterland of the TD, through a combination of partitioning (dealkalization/desalination, sterilization, and drying) and temperature-controlled experiments, we found that the contribution of various shifting sand components to the total CO 2 flux in the shifting sand is strongly linked to soil temperature. The soil temperature difference between depths of 0 and 10 cm and the rate of soil temperature change at a depth of 10 cm are the two main factors driving CO 2 exchange in the shifting sand. Besides, we obtained the contribution of the CO 2 fluxes of various components in the TD shifting sand (Fig. 1a) . Comparing demonstrate that the expansion/contraction of soil air containing CO 2 caused by heat fluctuation in shifting sand provides an unexpectedly strong contribution towards total CO 2 exchange. This long-hidden process in combination with salts/alkali
Long-Term Assessment of Rehabilitation Treatment of Sports through Artificial Intelligence Research
2021
Computational and Mathematical Methods in Medicine
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Background. Artificial intelligence (AI) technology has been incorporated into all walks of life, especially the integration of machine learning and health management has achieved very significant progress and results. It is very necessary to analyze personalized sports health management services and long-term assessment of health issues in the era of AI. Methods. This paper explores AI + personalized sports management service system design ideas, system operation process, management stage
doi:10.1155/2021/4980718
pmid:34976108
pmcid:PMC8716213
fatcat:bxeislomsngsdoujhmtqzoztgi
more »
... n, taking common chronic diseases, and diabetes as examples. 150 patients were divided into an observation group and a control group, and the blood glucose, blood pressure, blood lipid, and knowledge awareness rate were compared. Results. The blood glucose, blood pressure, and blood lipid levels of the observation group all reduced, and the awareness rate of diabetes knowledge increased, which proved that the AI research has great value in sports rehabilitation research coupled with long term health assessment and is worth further research. Conclusion. The AI research proposed in this paper is of far-reaching practical significance in helping the transformation and upgrading of the sports health management service industry, promoting the innovative development of sports health management service supply, and promoting national fitness and national health.
Amiloride Enhances Antigen Specific CTL by Faciliting HBV DNA Vaccine Entry into Cells
2012
PLoS ONE
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf.
The induction of relatively weak immunity by DNA vaccines in humans can be largely attributed to the low efficiency of transduction of somatic cells. Although formulation with liposomes has been shown to enhance DNA transduction of cultured cells, little, if any, effect is observed on the transduction of somatic tissues and cells. To improve the rate of transduction, DNA vaccine delivery by gene gun and the recently developed electroporation techniques have been employed. We report here that to
doi:10.1371/journal.pone.0033015
pmid:22438887
pmcid:PMC3306379
fatcat:eme44i3w4jcwll6ckljw2uvnfe
more »
... circumvent requirement for such equipment, amiloride, a drug that is prescribed for hypertension treatment, can accelerate plasmid entry into antigen presenting cells (APCs) both in vitro and in vivo. The combination induced APCs more dramatically in both maturation and cytokine secretion. Amiloride enhanced development of full CD8 cytolytic function including induction of high levels of antigen specific CTL and expression of IFN-c+perforin+granzymeB+ in CD8+ T cells. Thus, amiloride is a facilitator for DNA transduction into host cells which in turn enhances the efficiency of the immune responses.
Effects of Cropland Conversion and Climate Change on Agrosystem Carbon Balance of China's Dryland: A Typical Watershed Study
2018
Sustainability
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf.
the Knowledge Innovation Program of Institute of Soil Science, Chinese Academy of Sciences grant number ISSASIP1630; and the Startup Foundation for Introducing Talent of NUIST grant number 2017r036 to Chaofan ...
doi:10.3390/su10124508
fatcat:ulqxyqcgxfeabl7hdl2wy5jqhe
Reactivation of Denervated Schwann Cells by Embryonic Spinal Cord Neurons to Promote Axon Regeneration and Remyelination
2019
Stem Cells International
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf.
In peripheral nerve injuries (PNIs) in which proximal axons do not regenerate quickly enough, significant chronic degeneration of Schwann cells (SCs) can occur at the distal stump of the injured nerve and obstruct regeneration. Cell transplantation can delay the degeneration of SCs, but transplanted cells fail to generate voluntary electrical impulses without downstream signal stimulation from the central nervous system. In this study, we combined cell transplantation and nerve transfer
doi:10.1155/2019/7378594
pmid:31885623
pmcid:PMC6915008
fatcat:oposrdkoffgorizeg6dvwpihvq
more »
... es to investigate whether the transplantation of embryonic spinal cord cells could benefit the microenvironment of the distal stump of the injured nerve. The experiment consisted of two stages. In the first-stage surgery, common peroneal nerves were transected, and embryonic day 14 (E14) cells or cell culture medium was transplanted into the distal stump of the CPs. Six months after the first-stage surgery, the transplanted cells were removed, and the nerve segment distal to the transplanted site was used to bridge freshly cut tibial nerves to detect whether the cell-treated graft promoted axon growth. The phenotypic changes and the neurotrophic factor expression pattern of SCs distal to the transplanted site were detected at several time points after cell transplantation and excision. The results showed that at different times after transplantation, the cells could survive and generate neurons. Thus, the neurons play the role of proximal axons to prevent chronic degeneration and fibrosis of SCs. After excision of the transplanted cells, the SCs returned to their dedifferentiated phenotype and upregulated growth-associated gene expression. The ability of SCs to be activated again allowed a favorable microenvironment to be created and enhanced the regeneration and remyelination of proximal axons. Muscle reinnervation was also elevated. This transplantation strategy could provide a treatment option for complex neurological injuries in the clinic.
Ultrasensitive monitoring of DNA damage associated with free radicals exposure using dynamic carbon nanotubes bridged interdigitated electrode array
2020
Environment International
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Chaofan Yu: Validation, Writing -review & editing. Zipeng Liu: Resources, Visualization. Ying Wang: Supervision, Funding acquisition, Conceptualization. ...
doi:10.1016/j.envint.2020.105672
pmid:32248022
fatcat:fd2m2exp6bfc5l7w7arefcucna
« Previous
Showing results 1 — 15 out of 100 results