检索结果-南通市图书馆

Understanding adversarial attacks on observations in deep reinforcement learning

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第5期67卷 69-83页

作者： You QIAOBEN Chengyang YING Xinning ZHOU Hang su jun ZHU Bo ZHANG Department of Computer Science and Technology Beijing National Research Center for Information Science and Technology Tsinghua-Bosch Joint Center for Machine Learning Institute for Artificial Intelligence Tsinghua University Peng Cheng Laboratory

Deep reinforcement learning models are vulnerable to adversarial attacks that can decrease the cumulative expected reward of a victim by manipulating its observations. Despite the efficiency of previous optimization-based methods for generating adversarial noise in supervised learning, such methods might not achieve the lowest cumulative reward since they do not generally explore the environmental dynamics.Herein, a framework is provided to better understand the existing methods by reformulating the problem of adversarial attacks on reinforcement learning in the function space. The reformulation approach adopted herein generates an optimal adversary in the function space of targeted attacks, repelling them via a generic two-stage framework. In the first stage, a deceptive policy is trained by hacking the environment and discovering a set of trajectories routing to the lowest reward or the worst-case performance. Next, the adversary misleads the victim to imitate the deceptive policy by perturbing the observations. Compared to existing approaches, it is theoretically shown that our adversary is strong under an appropriate noise level. Extensive experiments demonstrate the superiority of the proposed method in terms of efficiency and effectiveness, achieving state-of-the-art performance in both Atari and MuJoCo environments.

关键词： deep learning reinforcement learning adversarial robustness adversarial attack

Transient receptor potential vanilloid 1 involved in the analgesic effects of total flavonoids extracted from Longxuejie(Resina Dracaenae Cochinchinensis)

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Journal of Traditional Chinese Medicine 2024年第3期44卷 437-447页

作者： MO Xiaoqiang CHEN Yating YIN Qian CHEN Haibo BAN Qiang LI jun CHEN su YAO Jinguang Department of Stomatology Affiliated Hospital of Youjiang Medical University for Nationalities College of General Practitioners Youjiang Medical University for Nationalities College of Biomedical Engineering South-Central Minzu UniversityKey Laboratory of Cognitive Science of State Ethnic Affairs CommissionHubei Key Laboratory of Medical Information Analysis and Tumor Diagnosis and Treatment College of Pharmacy South-Central Minzu University

OBJECTIVE: To evaluate the analgesic effects of total flavonoids of Longxuejie(Resina Dracaenae Cochinchinensis)(TFDB) and explore the possible analgesic mechanism associated with transient receptor potential vanilloid 1(TRPV1).METHODS: Whole-cell patch clamp technique was used to observe the effects of TFDB on capsaicin-induced TRPV1 currents. Rat experiments in vivo were used to observe the analgesic effects of TFDB. Western blot and immunofluorescence experiments were used to test the change of TRPV1 expression in DRG neurons induced by TFDB.REsuLTS: Results showed that TFDB inhibited capsaicin-induced TRPV1 receptor currents in acutely isolated dorsal root ganglion(DRG) neurons of rats and the half inhibitory concentration was(16.7 ± 1.6) mg/L.TFDB(2-20 mg/kg) showed analgesic activity in the phase Ⅱ of formalin test and(0.02-2 mg per paw)reduced capsaicin-induced licking times of rats. TFDB(20 mg/kg) was fully efficacious on complete Freund's adjuvant(CFA)-induced inflammatory thermal hyperalgesia and capsaicin could weaken the analgesic effects. The level of TRPV1 expressions of DRG neurons was also decreased in TFDB-treated CFA-inflammatory pain rats.CONCLUSION: All these results indicated that the analgesic effect of TFDB may contribute to their modulations on both function and expression of TRPV1 channels in DRG neurons.

关键词： Freund's adjuvant Longxuejie () capsaicin flavonoids formaldehyde inflammatory pain transient receptor potential vanilloid 1

P-and SV-wave dispersion and attenuation in saturated microcracked porous rock with aligned penny-shaped fractures

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Petroleum Science 2024年第1期21卷 143-161页

作者： Sheng-Qing Li Wen-Hao Wang Yuan-Da su jun-Xin Guo Xiao-Ming Tang School of Geosciences China University of Petroleum(East China)Qingdao 266580ShandongChina Laoshan National Laboratory Qingdao 266071ShangdongChina Guangdong Provincial Key Laboratory of Geophysical High-resolution Imaging Technology Southern University of Science and TechnologyShenzhen 518055GuangdongChina Department of Earth and Space Sciences Southern University of Science and TechnologyShenzhen 518055GuangdongChina

P-and SV-wave dispersion and attenuation have been extensively investigated in saturated poroelastic media with aligned fractures.However,there are few existing models that incorporate the multiple wave attenuation mechanisms from the microscopic scale to the macroscopic scale.Hence,in this work,we developed a unified model to incorporate the wave attenuation mechanisms at different scales,which includes the microscopic squirt flow between the microcracks and pores,the mesoscopic wave-induced fluid flow between fractures and background(FB-WIFF),and the macroscopic Biot's global flow and elastic scattering(ES)from the fractures.Using Tang's modified Biot's theory and the mixed-boundary conditions,we derived the exact frequency-dependent solutions of the scattering problem for a single penny-shaped fracture with oblique incident P-and SV-waves.We then developed theoretical models for a set of aligned fractures and randomly oriented fractures using the Foldy approximation.The results indicated that microcrack squirt flow considerably influences the dispersion and attenuation of P-and SV-wave velocities.The coupling effects of microcrack squirt flow with the FB-WIFF and ES of fractures cause much higher velocity dispersion and attenuation for P waves than for SV waves.Randomly oriented fractures substantially reduce the attenuation caused by the FB-WIFF and ES,particularly for the ES attenuation of SV waves.Through a comparison with existing models in the limiting cases and previous experimental measurements,we validated our model.

关键词： Aligned fractures P-and SV-wave Dispersion and attenuation Microcracked porous background FB-WIFF Elastic scattering Squirt flow

A continuous and long-term in-situ stress measuring method based on fiber optic. Part I: Theory of inverse differential strain analysis

在线全文

学校读者我要写书评

暂无评论

Petroleum Science 2024年第2期21卷 1171-1189页

作者： Kun-Peng Zhang Mian Chen Chang-jun Zhao su Wang Yong-Dong Fan State Key Laboratory of Petroleum Resources and Prospecting China University of PetroleumBeijing102249China MOE Key Laboratory of Petroleum Engineering in China University of Petroleum Beijing102249China Beijing International Center for Gas Hydrate School of Earth and Space SciencesPeking UniversityBeijing100871China

A method for in-situ stress measurement via fiber optics was proposed. The method utilizes the relationship between rock mass elastic parameters and in-situ stress. The approach offers the advantage of long-term stress measurements with high spatial resolution and frequency, significantly enhancing the ability to measure in-situ stress. The sensing casing, spirally wrapped with fiber optic, is cemented into the formation to establish a formation sensing nerve. Injecting fluid into the casing generates strain disturbance, establishing the relationship between rock mass properties and treatment pressure.Moreover, an optimization algorithm is established to invert the elastic parameters of formation via fiber optic strains. In the first part of this paper series, we established the theoretical basis for the inverse differential strain analysis method for in-situ stress measurement, which was subsequently verified using an analytical model. This paper is the fundamental basis for the inverse differential strain analysis method.

关键词： In-situ stress Fiber optic Orthotropic elastic Differential evolution ABAQUS

Intelligent geochemical interpretation of mass chromatograms:Based on convolution neural network

在线全文

学校读者我要写书评

暂无评论

Petroleum Science 2024年第2期21卷 752-764页

作者： Kai-Ming su jun-Gang Lu Jian Yu Zi-Xing Lu Shi-Jia Chen Hubei Key Laboratory of Petroleum Geochemistry and Environment Wuhan430100HubeiChina College of Resources and Environment Yangtze UniversityWuhan430100HubeiChina School of Geoscience and Technology Southwest Petroleum UniversityChengdu610500SichuanChina Changqing Oilfield Company PetroChinaXi©an710021ShaanxiChina

Gas chromatography-mass spectrometry(GC-MS)is an extremely important analytical technique that is widely used in organic geochemistry.It is the only approach to capture biomarker features of organic matter and provides the key evidence for oil-source correlation and thermal maturity determination.However,the conventional way of processing and interpreting the mass chromatogram is both timeconsuming and labor-intensive,which increases the research cost and restrains extensive applications of this method.To overcome this limitation,a correlation model is developed based on the convolution neural network(CNN)to link the mass chromatogram and biomarker features of samples from the Triassic Yanchang Formation,Ordos Basin,China.In this way,the mass chromatogram can be automatically interpreted.This research first performs dimensionality reduction for 15 biomarker parameters via the factor analysis and then quantifies the biomarker features using two indexes(i.e.MI and PMI)that represent the organic matter thermal maturity and parent material type,respectively.subsequently,training,interpretation,and validation are performed multiple times using different CNN models to optimize the model structure and hyper-parameter setting,with the mass chromatogram used as the input and the obtained MI and PMI values for supervision(label).The optimized model presents high accuracy in automatically interpreting the mass chromatogram,with R2values typically above 0.85 and0.80 for the thermal maturity and parent material interpretation results,respectively.The significance of this research is twofold:(i)developing an efficient technique for geochemical research;(ii)more importantly,demonstrating the potential of artificial intelligence in organic geochemistry and providing vital references for future related studies.

关键词： Organic geochemistry Biomarker Mass chromatographic analysis Automated interpretation Convolution neural network Machine learning

Chaiqin Chengqi Decoction as an adjuvant treatment for mild/moderately severe hypertriglyceridemic acute pancreatitis: A retrospective study

在线全文

学校读者我要写书评

暂无评论

World Journal of Clinical Cases 2024年第11期12卷 1918-1928页

作者： Hai-Fu Zhang Ze-Xuan su Yong-Hang Feng Shuo-jun Li Bi-Yun Xie Department of Internal Medicine The First People's Hospital of FuyangHangzhou 311400Zhejiang ProvinceChina

Hypertriglyceridemia is the third leading cause of acute pancreatitis(AP),and its incidence is increasing.Due to its relatively insidious etiology,it is easy to be ignored in the early stages.In China,Chaiqin Chengqi Decoction(CQCQD)has long been employed for treating AP.AIM To evaluate the effectiveness of CQCQD in patients diagnosed with mild/moderately severe hypertriglyceridemic AP(HTG-AP).METHODS In this study,the clinical data of 39 patients with HTG-AP admitted from January 2019 to November 2022 were collected.The changes of blood lipids,gastrointestinal symptoms,and abdominal pain before and after treatment were analyzed and compared between the two groups.REsuLTS Twenty patients were treated with the conventional HTG-AP regimen,and 19 patients were additionally treated with CQCQD.After receiving treatment,the triglycerides(TG)level of the CQCQD group was lower than that of the CQCQD group(3.14±0.25 mmol/L vs 4.96±0.47 mmol/L,P<0.01).After 3 d of treatment,the patients in the CQCQD group had more bowel movements than the control group(2.51±0.25 times vs 1.00±0.17 times,P=0.01).The gastrointestinal function of most patients returned to normal,and the acute gastrointestinal injury score was significantly lower than that of the control group(0.11±0.07 vs 0.42±0.11,P<0.01).CONCLUSION In patients with HTG-AP,CQCQD can significantly reduce the TG level,shorten the recovery time of defecation,significantly improve the gastrointestinal function.

关键词： Hypertriglyceridemic acute pancreatitis Chinese medicine Chaiqin Chengqi Decoction Serum lipid Triglycerides

Serrated flow behavior of GH536 superalloy under different loading rates at room temperature

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Rare Metals 2024年第5期43卷 2317-2323页

作者： Wen-Xiang Gao Dong Ma Li-Li Zhao Fan Ye Ling Shao su-jun Wu School of Materials Science and Technology Beihang University

Serrated flow phenomenon has been widely observed among different metals due to the potential influence on their applications.Uniaxial tensile tests of nickel-based superalloy GH536 were carried out at loading rates of 0.06,0.60,3.60,and 36.00 mm·min-1at room temperature,respectively.The tensile stress–strain curves demonstrate repetitive and discontinuous yielding behavior,namely serrated flow.It is observed that the stress–strain curves were dominated by type B serration,and original annealing twin boundaries （TBs） were distorted to various degrees under comparatively different lower tensile rates.In contrast,the TBs almost disappear,and the type B serrations become smoother and regular under the higher loading rates.This phenomenon can be attributed to the interactions between TBs and dislocations during tensile deformation.Coupled effects of the mobile and immobile dislocations illustrate the unsteady amplitude of serrations observed in stress–strain curves.Transmission electron microscope images of tested pieces reveal the interaction of dislocation and TB,with dislocation tangling around the TBs.

关键词： Nickel-based superalloy Serrations Twin boundaries Deformation mechanism

Targeted screening of an anti-inflammatory polypeptide from Rhopilema esculentum Kishinouye cnidoblasts and elucidation of its mechanism in alleviating ulcerative colitis based on an analysis of the gut microbiota and metabolites

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Food Science and Human Wellness 2024年第3期13卷 1336-1347页

作者： Ziyan Wang Qiuyue Shi Ying Feng Jiaojiao Han Chenyang Lu jun Zhou Zhonghua Wang Xiurong su State Key Laboratory for Managing Biotic and Chemical Threats to the Quality and Safety of Agro-products Ningbo UniversityNingbo 315000China School of Marine Science Ningbo UniversityNingbo 315000China Key Laboratory of Aquaculture Biotechnology(Ningbo University) Ministry of EducationNingbo UniversityNingbo 315000China College of Life Sciences Tonghua Normal UniversityTonghua 134000China Shandong Beiyou Biotechnology Co. Ltd.Weifang 261000China

Ulcerative colitis(UC)is a recurrent inflammatory bowel disease that imposes a severe burden on families and society.In recent years,exploiting the potential of marine bioactive peptides for the treatment of diseases has become a topic of intense research interest.This study revealed the mechanism underlying the protective effect of the dominant polypeptide PKKVV(Pro-Lys-Lys-Val-Val)of Rhopilema esculentum cnidoblasts against DSS-induced UC through a combined analysis of the metagenome and serum metabolome.Specifically,the polypeptide composition of R.esculentum cnidoblasts was determined by matrix-assisted laser desorption ionization time-of-flight mass spectrometry(MALDI-TOF/TOF-MS).Molecular docking showed that the dominant peptide PKKVV could bind better with tumor necrosis factor-α(TNF-α)than the original ligand.subsequent animal experiments suggested that PKKVV could modulate disorganized gut microorganisms in mice with UC;affect serum metabolites through the arachidonic acid,glycerophospholipid and linoleic acid metabolism pathways;and further alleviate UC symptoms.This study provides a reference for the comprehensive development of marine bioactive substances and nonpharmaceutical treatments for UC.

关键词： Ulcerative colitis Rhopilema esculentum Kishinouye Cnidoblasts Marine bioactive polypeptides Metagenome Serum metabolome

Improvement of the Bayesian neural network to study the photoneutron yield cross sections

在线全文

学校读者我要写书评

暂无评论

Nuclear Science and Techniques 2022年第11期33卷 1-9页

作者： Yong-Yi Li Fan Zhang jun su Department of Computing Changzhi UniversityChangzhi 046011China Department of Physics Changzhi UniversityChangzhi 046011China Sino-French Institute of Nuclear Engineering and Technology Sun Yat-sen UniversityZhuhai 519082China

This work is an attempt to improve the Bayesian neural network (BNN) for studying photoneutron yield cross sections as a function of the charge number Z, mass number A, and incident energy ε. The BNN was improved in terms of three aspects:numerical parameters, input layer, and network structure. First, by minimizing the deviations between the predictions and data, the numerical parameters, including the hidden layer number, hidden node number, and activation function, were selected. It was found that the BNN with three hidden layers, 10 hidden nodes, and sigmoid activation function provided the smallest deviations. Second, based on known knowledge,such as the isospin dependence and shape effect, the optimal ground-state properties were selected as input neurons. Third, the Lorentzian function was applied to map the hidden nodes to the output cross sections, and the empirical formula of the Lorentzian parameters was applied to link some of the input nodes to the output cross sections. It was found that the last two aspects improved the predictions and avoided overfitting, especially for the axially deformed nucleus.

关键词： Bayesian neural network Photoneutron cross sections Giant dipole resonance