ntry-header

 

CONFERENCE INFORMATION ORGANIZERS KEYNOTE SPEAKERS PROGRAM REGISTRATION LOCAL INFORMATION CONTACT US
KEY DATES
May 20th, 2018 11:00 AM Registration open
August 1st, 2018 10:00 AM Payment open
September 1st, 2018 05:00 PM Early registration deadline
September 30th, 2018 05:00 PM Registration & Cancellation deadline

Ⅰ. Conference Information

With the rapid development of modern computing technology and the rise of the big data, opportunities and challenges in statistical computing have been witnessed in all dimensions of data science. Confronted to such condition, the Asian Regional Section of the International Association for Statistical Computing (IASC-ARS) and Chinese Association for Statistical Computing (IASC) jointly hold the IASC-ARS 25th Anniversary Conference & CASC 2nd Annual Conference in Beijing, China. The conference is dedicated to promoting further development of statistical computing in modern data science, and to providing a platform for international academic exchanges and cooperation among Statistical Computing and data science professionals.

TimeNovember 9 – 11, 2018

Venue Beijing Conference Center, Beijing,China

(No. 88 Laiguangying West Road, Chaoyang District, Beijing)

Ⅱ. Organizers

The Asian Regional Section of the International Association for Statistical Computing (IASC-ARS)

The Asian Regional Section (ARS) of the International Association for Statistical Computing (IASC) is established by the Council in accordance with art. 12 of the IASC Statutes in 1993 by the name of the East Asian Regional Section (EARS) and the name is changed to the Asian Regional Section in 1998. It is composed of the members of the IASC residing in Asia. Our mission: to organize international or regional seminars, conferences, meetings. In particular, the ARS shall be responsible for arranging Asian meetings on computational statistics, which are abbreviated to IascAsian meetings hereafter; to promote research and training programs on theoretical and practical aspects of computational statistics; to foster evaluations of statistical computing techniques and programs; to publish scientific periodicals and books, reports and newsletters, independently or in cooperation with other organizations; to collaborate with other bodies having similar objectives.

Chinese Association for Statistical Computing (CASC)

The recent advancement of statistical research and increasing requirements arisen from the big data in various application fields have provided unprecedented challenges and opportunities in statistical computing. In order to respond to the pressing needs and seize the historical opportunity, a group of young statisticians, led by Professor Ke Deng of Center for Statistical Science at Tsinghua University, have jointly initiated the Chinese Association for Statistical Computing, as a branch of Chinese Association for Applied Statistics. The CASC was launched on March 19, 2017. The secretariat of CASC is located in the Center for Statistical Science at Tsinghua University.

CASC aims to unite the professionals in both academy and industry to enhance research and applications of statistical computing, towards boosting new theories and methodologies, promoting the applications in health, economics, industry, and other related fields, and advancing the development of statistics in China and overseas.

Co-organizer: Center for Statistical Science of Tsinghua University (TCSS)

On May 16, 2014, the Center for Statistical Sciences of Tsinghua University (TCSS) was officially launched by the university council. The Center will foster and coordinate the development of statistical science in Tsinghua University. The goal is to build a powerful faculty team, to develop cutting-edge research, to strengthen collaborations between disciplines, and, in short, to establish a world-class center of excellence in statistics at Tsinghua University. The Center is an independent academic unit at the school level, while for administrative affairs it is affiliated to the Department of Industrial Engineering.

The Center is dedicated to promoting research and teaching of statistics in Tsinghua. Research areas include mathematical statistics, computational statistics, biostatistics, with emphasis on interdisciplinary studies, especially between statistics and life sciences, medicine, engineering, and business. With their recognized research excellence, engineering, business, and life sciences at Tsinghua University are well placed to benefit from as well as provide nourishment to statistical science. By integrating research resources across campus, The Center has set its goal on high impact academic achievements in both theoretical and applied statistics.

Ⅲ. Keynote Speakers

Coming soon…

Ⅳ. Program

Coming soon…

Ⅴ. Registration

1.   Registration Date

Early Registration: Before September 1, 2018

Regular Registration: September 1-30, 2018

Registration deadline: September 30, 2018

2.   Registration Fee

Early Registration Regular Registration
Regular 960 RMB 1200 RMB
Student 600 RMB 750 RMB

Note: Luncheons and a banquet will be covered by the sponsor during the conference. All participants are responsible to their own accommodation and transportation.

Accommodation Fee

Standard Room at Building 6# Standard Room at Building 9#
Price 550 RMB 650 RMB

Your room contains a piece of breakfast. The accommodation fee will be charged when you check in.

3 Registration Procedure

Step 1: Please fill the registration information.

Step 2: Please click the link to charge the registration fee.

Registration fee will be charged by August.

(1)Online registration

哦!我们不能找到您的表单。

Please send an email to qianbao@mail.tsinghua.edu.cn before Oct. 1st if your schedule is changed.

哦!我们不能找到您的表单。

 (2)Payment:

E-currency payment: Master, Visa

Ⅵ. Contact Us

Secretariart for Chinese Association for Statistical Computing
Address: Weiqing Building 212, Tsinghua University, Beijing 100084, China

Tel: +86 (10)-62786091

Fax:+86 (10) -62783842

Email:stats@tsinghua.edu.cn

ⅥⅠ. Local Information

Venue: The Beijing Conference Center

Address: No. 88 Laiguangying West Road, Chaoyang District, Beijing

Tel: 86(10)-84901668

http://www.beijinghuiyizhongxin.com/encontact/176.html

http://www.google.cn/maps/place/Beijing+Conference+Center/@40.0105258,116.4077197,13.75z/data=!4m5!3m4!1s0x35f1aab720d5746b:0xad5ff040d3d33ade!8m2!3d40.021009!4d116.42843?hl=en

Local Transportation

Local transportation from the airport to the conference center

 

Hotel Information

Address: No. 88 Laiguangying West Road, Chaoyang District, Beijing

Tel: 86(10)-84901668

http://www.beijinghuiyizhongxin.com/encontact/176.html

Standard Room

Scenic Spots in Beijing

Tian’anmen (the Gate of Heavenly Peace)

Tian’anmen (the Gate of Heavenly Peace), is located in the center of Beijing. It was first built in 1417 and named Chengtianmen (the Gate of Heavenly Succession). At the end of the Ming Dynasty, it was seriously damaged by war. When it was rebuilt under the Qing in 1651, it was renamed Tian’anmen, and served as the main entrance to the Imperial City, the administrative and residential quarters for court officials and retainers. The southern sections of the Imperial City wall still stand on both sides of the Gate. The tower at the top of the gate is nine-room wide and five –room deep. According to the Book of Changes, the two numbers nine and five, when combined, symbolize the supreme status of a sovereign. During the Ming and Qing dynasties, Tian’anmen was the place where state ceremonies took place. On October 1, 1949, chairman Mao Zedong proclaimed on Tian’anmen Rostrum the founding of the People’s Republic of China. Since then Tian’anmen has been the symbol of New China.

 

The Great Wall of China

The Great Wall of China is a series of fortifications made of stone, brick, tamped earth, wood, and other materials, generally built along an east-to-west line across the historical northern borders of China to protect the Chinese states and empires against the raids and invasions of the various nomadic groups of the Eurasian Steppe. Several walls were being built as early as the 7th century BC; these, later joined together and made bigger and stronger, are collectively referred to as the Great Wall. Especially famous is the wall built in 220–206 BC by Qin Shi Huang, the first Emperor of China. Little of that wall remains. The Great Wall has been rebuilt, maintained, and enhanced over various dynasties; the majority of the existing wall is from the Ming Dynasty (1368–1644).

Apart from defense, other purposes of the Great Wall have included border controls, allowing the imposition of duties on goods transported along the Silk Road, regulation or encouragement of trade and the control of immigration and emigration.

The Great Wall stretches from Dandong in the east to Lop Lake in the west, along an arc that roughly delineates the southern edge of Inner Mongolia. A comprehensive archaeological survey, using advanced technologies, has concluded that the Ming walls measure 8,850 km (5,500 mi). This is made up of 6,259 km (3,889 mi) sections of actual wall, 359 km (223 mi) of trenches and 2,232 km (1,387 mi) of natural defensive barriers such as hills and rivers. Another archaeological survey found that the entire wall with all of its branches measures out to be 21,196 km (13,171 mi). Today, the Great Wall is generally recognized as one of the most impressive architectural feats in history.

The Temple of Heaven

The Temple of Heaven (Chinese: 天坛; pinyin: Tiantan) is an imperial complex of religious buildings situated in the southeastern part of central Beijing. The complex was visited by the Emperors of the Ming and Qing dynasties for annual ceremonies of prayer to Heaven for good harvest. It has been regarded as a Taoist temple, although Chinese heaven worship, especially by the reigning monarch of the day, predates Taoism.

In ancient China, the Emperor of China was regarded as the Son of Heaven, who administered earthly matters on behalf of, and representing, heavenly authority. To be seen to be showing respect to the source of his authority, in the form of sacrifices to heaven, was extremely important. The temple was built for these ceremonies, mostly comprising prayers for good harvests.

#post-11953
ntry-header

2018年5月15日,【统计学论坛】在清华大学伟清楼209成功举办。本次报告邀请到西南财经大学的蒋家坤博士。报告由清华大学统计学研究中心的杨立坚教授主持。本次报告的主题是“非高斯分布的函数型数据的聚类分析”。

蒋家坤博士
论坛现场
与会人员合影
#post-11951
ntry-header

2018年5月14日,【统计学论坛·特邀报告】在清华大学伟清楼209成功举办,本次报告邀请到清华大学数学科学系的丘成栋教授。报告由清华大学统计学研究中心副主任邓柯教授主持。本次报告的主题是“Protein universe and its applications”。当代数学大师、清华大学数学科学中心主任丘成桐先生亦亲临报告现场,并同现场师生针对学术问题展开讨论。

丘成栋教授
丘成桐先生
论坛现场
与会人员合影
#post-11950
ntry-header

会议通知

尊敬的各位同道:

由清华大学统计学研究中心、广州柏视医疗科技有限公司联合主办的智能医学影像分析研讨会将于2018年6月15日在北京清华紫光国际交流中心召开。

医学影像为医疗提供基础支撑,是临床最重要的诊断依据之一。由于医疗资源配置不均衡、培训的规范性及水平有待提高,基层影像诊断的误诊漏诊率偏高。人工智能技术的发展,尤其是深度学习等新技术在图像和文本分析等领域的融合应用,推动了医学人工智能的快速发展。人工智能通过和医学技术的结合,赋能影像医学及肿瘤诊疗, 可减少大量简单重复性劳动, 辅助医生提升工作效率。

本次大会将深入探讨智能医学影像分析的热点问题,邀请国内外医学影像和人工智能领域的顶尖专家和企业代表,共同交流在医学影像中应用人工智能的思想,方法和实践,探讨行业趋势、典型案例,和技术动态,以期建立医学影像专家和人工智能专家间的深度战略合作,促进人工智能和医学影像的互联互通,提升智能医学影像分析研究的国际影响力。

本次大会不收注册费。衷心希望全国医学影像、人工智能的同行共同参与这一学宴,在此谨代表本次大会组委诚挚地邀请您并热切期待您的参与!

 

会议联系:

田女士  清华大学统计学研究中心     010-62786091

刘女士  广州柏视医疗科技有限公司  15300099673

组委会邮箱:stats@tsinghua.edu.cn

 

#post-11948
ntry-header

2018年5月8日,美国Suffolk大学商学院大学生交流访问团到访清华大学统计学研究中心。中心副主任邓柯教授以及杨立坚教授接待访问。几位教授分别从校园生活、学术研究、职业发展等几个方面介绍中心情况,并同来访学生亲切交流。随后我中心研究生姜瑛恺、杨洋、汪子栋、林毓聪作为学生代表向来访学生介绍统计学科的知识体系以及实际应用,并带领访问团学生参观清华校园,体验校园生活。

#post-11946
ntry-header

2018年5月7日,荷兰特温特大学地理信息学院助理教授贾鹏访问我中心,并做学术报告:地理遥感信息技术在公共卫生领域的应用。

贾鹏教授与中心教员座谈
贾鹏教授
论坛现场
与会人员合影
#post-11944
ntry-header

#post-11939
ntry-header

2018年4月23日,美国托雷多大学数学与统计系的刘嵘教授访问我中心,并做学术报告。报告题目是Empirical likelihood inference for generalized additive model。

杨立坚教授(左)与刘嵘教授
刘嵘教授
论坛现场
与会人员合影
#post-11938
ntry-header

2018年4月16日,【统计学论坛·特邀报告】在清华大学主楼接待厅成功举办。报告的主讲人是美国国家科学院院士、斯坦福大学统计系王永雄教授。作为清华大学107周年校庆的系列活动之一,王永雄教授的特邀报告从信息发布伊始即受到校内师生的广泛关注,活动现场更是气氛热烈,座无虚席。

报告现场

本次报告的主办单位为清华大学工业工程系统计学研究中心,并由统计学研究中心副主任邓柯副教授担任主持人。报告的题目是Mini-batching in Markov Chain Monte Carlo Inference。

王永雄 教授

王教授首先介绍问题背景。即在抽样问题中,如果目标分布不能很容易地直接抽样,用Markov Chain Monte Carlo (以下简称MCMC) 方法可以有效地解决这个问题。构造一条马氏链,使得它的极限分布是目标分布,迭代多次之后可以近似地得到目标分布的样本。Metropolis-Hastings (MH) 算法是 MCMC 方法中最常用的算法之一。当我们要抽参数的后验分布时,M-H 算法计算 M-H ratio 的时候需要用到全部数据。但是当数据量很大时,这个方法就不会有很高的效率。如果用 mini-batch 的方法,每次只需要计算少量数据的信息,可以有效地提高效率。由此引入了报告主要研究的问题 mini-batch tempered MCMC。

实现这个方法运用了统计学里经典的参数扩张 (parameter augmentation) 的办法。通过构造一个比参数空间更高维的分布,并在这个分布上运用mini-batch M-H 算法,抽到的样本取边缘分布,即可得到我们想要的参数的样本。通过数学方法可以证明,得到样本的极限分布是参数后验分布的一个回火版本 (tempered version)。随机模拟和实际数据两个例子表明,MINi-batch Tempered MCMC (MINT) 算法能得到和 Tempered MCMC 方法类似的样本数据,但是计算和用时方面有显著地提高。

 

第二个问题是由于 MINT 算法抽样得到的样本并不是真的后验分布,只是后验分布升温后的一个版本。如果我们想要得到真实的后验分布的样本,王教授介绍了 equi-energy (EE) sampler的办法(Kou, Zhou and Wong, 2006)。原来的EE方法可以有效地解决分布多峰(multi-mode)的问题,但是对后验分布抽样时,MCMC方法每一步还是需要所有数据的信息。把 MINT 和 EE 方法结合起来,可以得到 MINi-batch Tempered Equi-Energy (MINTEE) 算法。MINT可以有效地从高温分布中抽样;EE方法要求从最高的温度开始,在每个温度中都抽相应的马氏链,并不断更新EE set,最低温分布就是目标分布,可以得到想要的样本。同样地,MINTEE 方法在实际应用中有很高地效率,计算复杂度也比EE方法小。

 

第三部分王教授介绍了 cone move 的方法。在机器学习里面流行的Langevin Dynamics方法,每次马氏链更新的时候,proposal distribution 的期望会沿着梯度方向,而这种方法会使 proposal distribution 的反向概率 (reverse probability) 很小,从而导致 M-H ratio 也非常小,马氏链很难转移出去,抽样效率低。王教授介绍的方法是提出一个新的proposal distribution,空间上像是把两个圆锥反向按顶点接在一起,它在正负梯度上有着相同的概率,这样每次马氏链的转移是可逆的(reversible),效率会更高一些。

 

最后的提问环节,老师和同学们讨论了和 MINT 方法相关的问题:能否把 MINT 方法运用在 Gibbs 抽样的框架之下?MINTEE 方法中,不同温度下的马氏链可不可以在不同温度之间进行转移等等。王教授回答到,在Gibbs抽样中运用mini-batch方法是十分有趣也十分有挑战的问题,期望看到更多的相关的进展。对于后一个问题,在parallel tempering方法中,多条马氏链可以在不同温度之间交换,而在EE sampler方法里,高温的马氏链可以帮助指导或影响低温的马氏链,低温对高温是不会有影响的。

 

世界范围内统计学科顶尖院士的特邀报告是统计学论坛的传统活动,每一学年均如期举办,它为清华的师生提供了一个和学术大咖近距离沟通交流的机会,获得一致好评与称赞。

现场合影
#post-11934
ntry-header

2018年4月14日,医疗大数据与卫生技术评估论坛在文津国际酒店举行,本次论坛由清华大学统计学研究中心医疗大数据中心和国家卫生健康委员会卫生技术评估重点实验室(复旦大学)联合主办。清华大学统计学研究中心的战略合作伙伴——Analysis Group,作为协办单位为本次论坛提供了大力支持。哈佛大学生物统计系系主任林希虹教授,人力资源和社会保障部社会保险事业管理中心熊先军书记,美国国家科学院院士、斯坦福大学王永雄教授,北京大学刘国恩教授,清华大学邓柯副教授,Analysis Group吴琼博士,复旦大学陈英耀教授,北京大学人民医院王杉教授,哈尔滨工业大学王亚东教授课题组刘博副教授,中电数据服务有限公司首席应用官肖辉先生,英国谢菲尔德大学Praveen Thokala教授,清华大学张学工教授,Analysis Group韩思蒙博士,杜兰大学施李正教授,天津大学吴晶教授,国家食品安全风险评估中心风险监测部肖革新副主任等医疗领域学界和业界人士参加了本次论坛,并展开了热烈的讨论。

论坛现场

首先,清华大学统计学研究中心副主任邓柯副教授作为主办方负责人介绍到场嘉宾,并向本次论坛的组织方和出席嘉宾致谢。

林希虹 教授

随着哈佛大学生物统计系系主任林希虹教授的开幕致辞,论坛正式拉开帷幕,林教授列举了数据科学鼎盛时代,医疗领域的几大方向,如人类基因组计划、电子病历数据等,充满挑战和机遇,希望我们能迎难而上,共同将人类医疗向前迈一大步。

熊先军 书记

作为本次论坛正式报告的首位出场专家,人力资源和社会保障部社会保险事业管理中心熊先军书记,根据30年的从业经验,从卫生经济学(PE)是什么切入,论述了卫生经济学在政府决策中的重要作用。

王永雄 教授

随后,美国国家科学院院士、斯坦福大学王永雄教授对电子病历中的精准医学的研究谈了几点自己的思考。从电子病历中抓取的临床数据和人类基因组测序的基因数据有机结合,是未来精准医疗的突破点,国内目前亟需加大财政投入力度,建设人群基因库。

刘国恩 教授

北京大学刘国恩教授从中国2010年之后经济增长缓慢的现象说起,分享了不同学者对经济形势的看法,进而从经济学角度肯定了医疗在未来经济学中的重要地位。

邓柯 副教授

清华大学邓柯副教授则从医学自然语言处理的统计学方法为切入点,介绍了具体的无监督的文本分析方法在电子病历中的成功应用,带大家领略到统计模型在处理医疗电子病历中的重要潜力。

吴琼 博士

业界专家Analysis Group的吴琼博士,介绍了在卫生技术评估中识别最优人群的重要性和具体实现方式,寻找满足最少标准的最大目标人群,对临床试验、决策制定、医疗服务都是非常重要的。

陈英耀 教授

复旦大学陈英耀教授从中国卫生技术评估的发展现状和展望展开设想,从卫生技术面临的挑战和引导案例出发,阐述了卫生技术评估需要综合考虑诸多因素,有效性评价、伦理评价、评估和评审等,最后提出希望未来厘清各方利益关系,通过需求倒逼HTA发展的展望。

王杉 教授

北京大学人民医院的王杉教授作为论坛下午的第一位演讲嘉宾,从实际应用出发,阐明了大数据在医疗领域的应用场景,如达芬奇机器人、智慧人工器官、智慧医政等,并从数据共享和如何有效提取数据方面介绍中国了医疗数据带来的特殊性和挑战性。

张学工 教授

随后,清华大学的张学工教授对中国医疗面临的挑战进行了补充,根据自身的切实体会阐述了中国医疗不够精准和医疗不够普惠的问题,并提出一些建议,希望能改革医疗数据管理制度,从建立健全“个人大数据”和“全民大数据”方面入手,解决医疗大数据共享的法律困境。

肖辉 先生

中电数据服务有限公司首席应用官肖辉先生,首先介绍中电数据的发展现状,主题是大数据、大健康、大发展,围绕惠民、惠企、惠政这一总体规划,数据安全这一核心,希望能通筹建国家公司、地方分公司以及与国外公司合作,促进产业的发展。

Praveen Thokala 教授

来自英国谢菲尔德大学的Praveen Thokala教授则分享了英国的健康技术评估的制度和方法,对我国相关领域的研究和发展有重要的借鉴意义。

刘博 副教授

哈尔滨工业大学王亚东教授课题组的刘博副教授陈述了目前我国基因组研究的现状,分析了大规模基因组快速比对的技术挑战,并介绍了在相关领域的最新成果。

韩思蒙 博士

最后的演讲者是来自Analysis Group的韩思蒙博士,韩博士介绍了医疗大数据在真实世界的应用场景,通过大量案例的分析,总结出了医疗大数据、统计学方法和计算平台的支持是证据产生的重要手段。

圆桌讨论
精彩纷呈的圆桌讨论环节使论坛的气氛达到高潮。参与嘉宾分别是王永雄教授、施李正教授、吴晶教授、肖革新副主任以及肖辉先生,邓柯副教授担任主持人,嘉宾们围绕中国医疗大数据的发展的挑战和机遇问题进行了深入探讨。王永雄教授认为产业和政府需要增强合作,将基因组数据整合起来;吴晶教授提到应将医保融入医疗大数据,加强中外医疗的合作;施李正教授则认为要从政府对人才的需求和提高研究质量方面解决目前的问题;肖辉提到要解决技术方面存在的问题、完善法规和相关体系;肖革新副主任希望从数据质量、数据交流、机制创新、政府决策需要、老百姓和企业的需求方面,提升未来的供给制改革。
       论坛在陈英耀教授的闭幕词中圆满结束。陈教授提到,这次医疗大数据与卫生技术评估论坛是一个多学科多领域参与的盛会,为国际和国内专家交流合作提供了很好的机会,希望这个论坛能继续发展并延续,共同推进医疗和卫生领域的发展。

与会人员合影

 

媒体报道:http://finance.ifeng.com/a/20180420/16161350_0.shtml

(来源:凤凰网)

#post-11933