PKU

 Statistics and Information Techonlogy
 Summer School 2010
 Peking University


Course Information

Organizer

微软 Statistics & Information Technology Laboratory of Peking University
Beijing International Center for Mathematics Research, Peking University
Statistical Center at Peking University

Courses

Diane Lambert, 谷歌 Inc.
June 15-24, An Intense Course in Data Analysis Using Multi-Level Regression Models,
Rm 425, Ying Jie Exchange Center, PKU
June 21, 2-4PM, seminar: Quality Control at 谷歌 Scale
Rm 425, Ying Jie Exchange Center, PKU

Course Description [pdf]: Data is the raw material of knowledge, and computing, graphics, and statistical models are the tools that statisticians use to extract information from data. This course will expand on that theme through a series of lectures and team laboratory projects related to building, interpreting and validating regression models, especially those with binary responses or multiple levels of randomness.
In addition to the course, Dr. Lambert will give a seminar titled `Quality Control at 谷歌 Scale' that will be open to anyone. Loosely speaking, quality controls usually implies continu- ously improving industrial products by experimenting with small, local changes in engineering and management processes. At the core of quality improvement lie measurement, experimentation, and learning followed by implementation. This talk will show how 谷歌 uses these well-established quality principles (along with huge amounts of data) to improve seach and ads for users, advertisers, and publishers.
Course Reader: Multilevel Models, by Gelman and Hill, 2007

Tentative Schedule:
DateLectures (Ying Jie Exchange Center #425)SlidesCodeDataAssignments
Tue June 15 linear regression
10-12pm lecture, 2-4pm lab
Lecture1.pdf Lecture1.Rsleep1.csv Assignment1.pdf
Wed June 16more on linear regression
10-12pm lecture, 2-4pm lab
Lecture2.pdf Lecture2.Rsleep1.csv Assignment2.pdf
Thu June 17logistic regression
10-12pm lecture
Lecture3_Assignments12.pdf regAssignments.Rsleep1.csv
Mon June 212-4pm, seminar on Quality Control at 谷歌 Scale
Tue June 22logistic regression
10-12pm lecture, 2-4pm lab
Lecture4_glm1.pdf logReg1.Rwells.csv
Wed June 23multilevel linear models
10-12pm lecture, 2-4pm lab
Assignment3.pdf
Thu June 24multilevel logistc regression
10-12pm lecture, 2-4pm lab
Lecture5_multi2.pdf radon.csvmultiLevelAssignment.pdf


David Pollard, Department of Statistics and Mathematics, Yale University
July 13-23, A Very Short Course on Le Cam Theory
Course Web: http://www.stat.yale.edu/~pollard/Beijing2010/

Peter Bartlett, Department of Statistics and Division of Computer Science, UC Berkeley
July 13-16, Statistical learning methods for online and probabilistic prediction problems
Course Web: http://www.stat.berkeley.edu/~bartlett/talks/BeijingCourse2010.html
Course Description : This short course will provide an introduction to the design and theoretical analysis of prediction methods, focusing on statistical and computational aspects. It will cover probabilistic and game theoretic formulations of prediction problems, and will examine questions about the guarantees we can prove about the performance of prediction algorithms and the inherent difficulty of prediction problems. (Lectures in 3 days, 2 hours per day.)

Combined Schedule during 7/13-23 :
TimePlaceTalks
Tue July 13
10am-12pm
1:30-3:30pm
4-5pm

#1 Sci Bldg, Rm 1560 (理科一号楼 1560)
#1 Sci Bldg, Rm 1560 (理科一号楼 1560)
#1 Sci Bldg, Rm 1114 (理科一号楼 1114)

David Pollard Lecture 1
Peter Bartlett Lecture 1
Tze Leung Lai (Stanford) Seminar, The 1st Pao-Lu Hsu Lecture at Peking University.
Wed July 14
10am-11am
11am-12pm
2-3pm
4-5pm

#1 Sci Bldg, Rm 1114
#1 Sci Bldg, Rm 1114
Lijiao Bldg 103 (理教 103)
#1 Sci Bldg, Rm 1560

Francis Bach (INRIA-ENS) Seminar, Sparse Hierarchical Dictionary Learning.
Tao Shi (OSU) Seminar, Multi-Sample Data Spectroscopic Clustering of Large Datasets using Nystrom Extension.
Peter Bartlett Seminar, Optimal Online Prediction in Adversarial Environments.
David Pollard Informal Session 1
Thu July 15
10am-12pm
3-4pm
4-5pm

#1 Sci Bldg, Rm 1560 (理科一号楼 1560)

David Pollard Lecture 2
David Pollard Informal Session 2
Harry Zhou (Yale) Seminar, Optimal Estimation of Large Covariance Matrices.
Fri July 16
10am-12pm
2-4pm

#1 Sci Bldg, Rm 1560

Peter Bartlett Lecture 2
Peter Bartlett Lecture 3
Mon July 19
10am-11am
3-4pm

#1 Sci Bldg, Rm 1560

David Pollard Lecture 4
David Pollard Informal Session 4
Wed July 21
10am-12pm

New Guanghua Bldg Rm 217 (光华管理学院新楼217教室)

Peter Hall (U Melbourne) Seminar: CONCEPT OF DENSITY FOR FUNCTIONAL DATA.
Thu July 22
10am-12pm
3-4pm

#1 Sci Bldg, Rm 1560

David Pollard Lecture 4
David Pollard Informal Session 4
Fri July 23
10am-12pm
3-4pm

#1 Sci Bldg, Rm 1560

David Pollard Lecture 5
David Pollard Informal Session 5

Registration

Diane Lambert's course will be held in a computer room with fifty computers equipped. Hence this course can open to a maximum of 50 participants. The participants may bring their own laptops. Experience in R language is preferred.

There are No seat-limitations for the other two courses.

Registration is closed. Please refer to the participant namelist.

Course Assistants

CHENG, Xiaoxing Daniel cheng_xiaoxing@126.com, Peking University and Brown University (to-be).
LI, Tianxi tianxilicb@gmail.com, Zhejiang University and Stanford University (to-be).
QU, Jianghan Jenny jianghan294@hotmail.com, Peking University and University of South California (to-be).
ZHANG, Biyuan Elise zhang.elise08@gmail.com, Peking University.
ZHAO, Die Zoey zhaodiecool@gmail.com, Peking University.





Back to Home
by YAO, Yuan.
XML 地图 | Sitemap 地图