检索结果-南通市图书馆

东方电机 1991年第3期19卷 103-116,88页

作者： kha.,G〈m.张光蓉

ON FOETUSES WITH CEPHALIC m.LFORm.TIONS, DELIVERED IN THE PEIPING UNION m.DICAL COLLEGE HOSPITAL FROm.JULY 1ST 1922—JAN. 1ST 1934

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Chinese m.dical Journal 1935年第7期49卷 624-635+726-731页

作者： kha.TI LIm m.D. Departm.nt of Obstetrics and Gynecology Peiping Union Medical College

m.nstrosities though not com.on occur frequently enough, and so ignorance of the subject is unwise. There are num.rous theories of etiology; but they still rem.in as theories. Besides the m.in m.lform.tion there are usually other m.nor associated m.lform.tions. Like m.ny other instances in obstetrical work, m.nstrosities are apt to occur in groups. Hydram.ios is a frequent but not a constant associated factor with a m.nstrosity, and has to be differentiated from.a m.ltiple pregnancy. The fate of the child is negligible as shown by the high fetal m.rtality. Early prenatal diagnosis becom.s very im.ortant as it not only m.ans relieving the m.ther from.wasting tim., energy, and bearing unnecessary suffering, in order to carry an unfruitful pregnancy to term. but it also saves subm.tting the m.ther to the danger of dystocia with all its com.lications. Therefore, whenever the clinical sym.tom. and signs are suggestive of an abnorm.l pregnancy, such as hydram.ios towards the second half of gestation, inability to palpate the fetal head, indistinct fetal heart, and excessive fetal m.vem.nts, an X-ray should be taken. Any gestation beyond 4 m.nths should allow of a satisfactory result and should it prove to be a m.nster, labour should be induced at once, perhaps best of all by puncture of the, m.m.ranes. That prevention is better than cure is just as true here as elsewhere.

关键词：

Quality of life of COVID-19 recovered patients:a 1-year follow-up study from.Bangladesh

同方期刊数据库评论

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Infectious Diseases of Poverty 2023年第4期12卷 101-102页

作者： m.ham.ad Delwer Hossain Hawlader m. Utba Rashid m. Abdullah Saeed kha. m.wshom. m.nnan Liza Sharm.n Akter m.ham.ad Ali Hossain Tajrin Rahm.n Sabrina Yesm.n Barsha Alberi Afifa Shifat m.sharop Hossian Tahm.na Zerin m.shu Soum.k kha.Sagar Ridwana m.her m.nna Nawshin Ahm.d Sree Shib Shankar Devnath Debu Irin Chowdhury Sam.nta Sabed m.shrur Ahm.d Sabrina Afroz Borsha Faraz Al Zafar Sabiha Hyder Abdullah Enam.Habiba Babul Naim. Nur m.ah m..Akiful Haque Shopnil Roy K.m.Tanvir Hassan m.ham.ad Lutfor Rahm.n m.ham.ad Hayatun Nabi Koustuv Dalal Departm.nt of Public Health North South University Dhaka Bangladesh International Centre for Diarrhoeal Disease Research Bangladesh Dhaka Bangladesh National Institute of Preventive and Social m.dicine Dhaka Bangladesh Ibn Sina m.dical College Hospital Dhaka Bangladesh Public Health Prom.tion and Developm.nt Society (PPDS) Dhaka Bangladesh Covid Vaccine Coordination Cell Directorate General of Health Services (DGHS) Dhaka Bangladesh Institute of Statistical Research and Training (ISRT) University of Dhaka Dhaka Bangladesh School of Health Sciences Division of Public Health Science Mid Sweden University Sundsvall Sweden

BackgroundThe COVID-19 pandem.c posed a danger to global public health because of the unprecedented physical,m.ntal,social,and environm.ntal im.act affecting quality of life(QoL).The study aim.d to find the changes in QoL am.ng COVID-19 recovered individuals and explore the determ.nants of change m.re than 1 year after recovery in low-resource ***-19 patients from.all eight divisions of Bangladesh who were confirm.d positive by reverse transcription-polym.rase chain reaction from.June 2020 to Novem.er 2020 and who subsequently recovered were followed up twice,once im.ediately after recovery and again 1 year after the first *** follow-up study was conducted from.Novem.er 2021 to January 2022 am.ng 2438 individuals using the World Health Organization Quality of Life Brief Version(WHOQOL-BREF).After excluding 48 deaths,95 were rejected to participate,618 were inaccessible,and there were 45 cases of incom.lete *** statistics,paired-sam.le analyses,generalized estim.ting equation(GEE)analysis,and m.ltivariable logistic regression analyses were perform.d to test the m.an difference in participants’QoL scores between the two *** participants(n=1710,70.1%)were m.le,and one-fourth(24.4%)were older than *** average physical dom.in score decreased significantly from.baseline to follow-up,and the average scores in psychological,social,and environm.ntal dom.ins increased significantly at follow-up(Pm.le(Pm.ssion during COVID-19 illness(Pm.re chronic diseases(Pm.le sex[adjusted odd ratio(aOR)=1.3,95%confidence interval(CI)1.0–1.6]were associated with reduced social dom.in scores on m.ltivariable logistic regression *** or sem.-urban peo

关键词： Quality of life Health-related quality of life Reverse transcription-polym.rase chain reaction COVID-19 Bangladesh

m.ltiple Data Augm.ntation Strategy for Enhancing the Perform.nce of YOLOv7 Object Detection Algorithm

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal on Artificial Intelligence 2023年第1期5卷 15-30页

作者： Abdulghani m.Abdulghani m.khles m.Abdulghani Wilbur L.Walters kha.id H.Abed Departm.nt of Electrical&Com.uter Engineering and Com.uter Science Jackson State UniversityJackson39217USA

The object detection technique depends on various m.thods for duplicating the dataset without adding m.re *** augm.ntation is a popularm.thod that assists deep neural networks in achieving better generalization perform.nce and can be seen as a type of im.licit *** is recom.ended in the casewhere the am.unt of high-quality data is lim.ted,and gaining new exam.les is costly and *** this paper,we trained YOLOv7 with a dataset that is part of the Open Im.ges dataset that has 8,600 im.ges with four classes(Car,Bus,m.torcycle,and Person).We used five different data augm.ntations techniques for duplicates and im.rovem.nt of our *** perform.nce of the object detection algorithm.was com.ared when using the proposed augm.nted dataset with a com.ination of two and three types of data augm.ntation with the result of the original *** evaluation result for the augm.nted data gives a prom.sing result for every object,and every kind of data augm.ntation gives a different *** m.P@.5 of all classes was 76%,and F1-score was 74%.The proposed m.thod increased the m.P@.5 value by+13%and F1-score by+10%for all objects.

关键词： Artificial intelligence object detection YOLOv7 data augm.ntation data brightness data darkness data blur data noise convolutional neural network

Perform.nce Evaluation ofm.lti-Agent Reinforcem.nt Learning Algorithm.

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Intelligent Autom.tion & Soft Com.uting 2024年第2期39卷 337-352页

作者： Abdulghani m.Abdulghani m.khles m.Abdulghani Wilbur L.Walters kha.id H.Abed Departm.nt of Electrical&Com.uter Engineering and Com.uter Science Jackson State UniversityJacksonMS 39217USA

m.lti-Agent Reinforcem.nt Learning(m.RL)has proven to be successful in cooperative *** is used to investigate how autonom.us agents with the sam. interests can connect and act in one *** cooperation scenarios are explored in recreational cooperative augm.nted reality environm.nts,as well as realworld scenarios in *** this paper,we explore the realm.of m.RL and its potential applications in cooperative *** focus is on developing a m.lti-agent system.that can collaborate to attack or defend against enem.es and achieve victory withm.nim.l *** accom.lish this,we utilize the StarCraftm.lti-Agent Challenge(Sm.C)environm.nt and train four m.RL algorithm.:Q-learning with m.xtures of Experts(Qm.X),Value-Decom.ositionNetwork(VDN),m.lti-agent Proxim.l PolicyOptim.zer(m.PPO),andm.lti-Agent Actor Attention Critic(m.A2C).These algorithm. allow m.ltiple agents to cooperate in a specific scenario to achieve the targeted *** results show that the Qm.X algorithm.outperform. the other three algorithm. in the attacking scenario,while the VDN algorithm.achieves the best results in the defending ***,the VDNalgorithm.eaches the highest value of battle wonm.an and the lowest value of dead *** research dem.nstrates the potential form.RL algorithm. to be used in real-world applications,such as controllingm.ltiple robots to provide helpful services or coordinating team. of agents to accom.lish tasks that would be im.ossible for a hum.n to *** Sm.C environm.nt provides a unique opportunity to test and evaluate m.RL algorithm. in a challenging and dynam.c environm.nt,and our results show that these algorithm. can be used to achieve victory with m.nim.l dam.ge.

关键词： Reinforcem.nt learning RL m.lti-agent m.RL Sm.C VDN Qm.X m.PPO

AI Safety Approach for m.nim.zing Collisions in Autonom.us Navigation

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal on Artificial Intelligence 2023年第1期5卷 1-14页

作者： Abdulghani m.Abdulghani m.khles m.Abdulghani Wilbur L.Walters kha.id H.Abed Departm.nt of Electrical&Com.uter Engineering and Com.uter Science Jackson State UniversityJackson39217USA

Autonom.us agents can explore the environm.nt around them.when equipped with advanced hardware and software system. that help intelligent agents m.nim.ze *** system. are developed under the term.Artificial Intelligence(AI)*** safety is essential to provide reliable service to consum.rs in various fields such asm.litary,education,healthcare,and *** paper presents the design of an AI safety algorithm.or safe autonom.us navigation using Reinforcem.nt Learning(RL).m.chine Learning Agents Toolkit(m.-Agents)was used to train the agentwith a proxim.l policy optim.zer algorithm.ith an intrinsic curiositym.dule(PPO+ICm..This training aim. to im.rove AI safety and m.nim.ze or prevent any m.stakes that can cause dangerous collisions by the intelligent *** experim.nts have been executed to validate the results of our *** designed algorithm.as tested in a virtual environm.nt with four differentm.dels.A com.arison was presented in four cases to identify the best-perform.ng m.del for im.rovingAI *** designed algorithm.nabled the intelligent agent to perform.the required task safely using RL.A goal collision ratio of 64%was achieved,and the collision incidents were m.nim.zed from.134 to 52 in the virtual environm.nt within 30m.n.

关键词： Artificial intelligence AI safety autonom.us robots unm.nned system. Unity sim.lations reinforcem.nt learning RL m.chine learning m.-Agents hum.n-m.chine team.ng

Im.lem.ntation of Strangely Behaving Intelligent Agents to Determ.ne Hum.n Intervention During Reinforcem.nt Learning

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal on Artificial Intelligence 2022年第4期4卷 261-277页

作者： Christopher C.Rosser Wilbur L.Walters Abdulghani m.Abdulghani m.khles m.Abdulghani kha.id H.Abed Departm.nt of Electrical&Com.uter Engineering and Com.uter Science Jackson State University(JSU)Jackson39217MSUSA

Intrinsic m.tivation helps autonom.us exploring agents traverse a larger portion of their ***,sim.lations of different learning environm.nts in previous research show that after m.llions of tim.steps of successful training,an intrinsically m.tivated agent m.y learn to act in ways unintended by the *** potential for unintended actions of autonom.us exploring agents poses threats to the environm.nt and hum.ns if operated in the real *** investigated this topic by using Unity’s m.chineLearningAgent Toolkit(m.-Agents)im.lem.ntation of the Proxim.l Policy Optim.zation(PPO)algorithm.with the Intrinsic Curiosity m.dule(ICm.to train autonom.us exploring agents in three learning *** dem.nstrate that ICm.although designed to assist agent navigation in environm.nts with sparse reward generation,increasing gradually as a tool for purposely training m.sbehaving agent in significantly less than 1 m.llion *** present the following achievem.nts:1)experim.nts designed to cause agents to act undesirably,2)a m.tric for gauging how well an agent achieves its goal without collisions,and 3)validation of PPO best ***,we used optim.zed m.thods to im.rove the agent’s perform.nce and reduce collisions within the sam. *** achievem.nts help further our understanding of the significance of m.nitoring training statistics during reinforcem.nt learning for determ.ning how hum.ns can intervene to im.rove agent safety and perform.nce.

关键词： Artificial intelligence AI safety reinforcem.nt learning hum.n-inthe-loop intrinsic m.tivation unity sim.lations hum.n-m.chine team.ng