纪念。

Reminiscences of a Statistician: The Company I Kept

by Erich Lehmann

许宝騄
数学家。在中国开创了概率论、数理统计的教学与研究工作。在内曼－皮尔逊理论、参数估计理论、多元分析、极限理论等方面取得卓越成就，是多元统计分析学科的开拓者之一。
在1945年秋，他应邀去美国加州大学伯克利分校和哥伦比亚大学任访问教授，各讲一个学期，学生中有安德森，莱曼等人。1946年到北卡罗莱纳大学任教。一年后，他决心回国，谢绝了一些大学的聘任，回到北京大学任教授。1948年他当选为中央研究院院士。

端木同学在阳光明媚的加州度过愉快的一周，并确认了两件很重要的事情。其中之一就是决定成为“第三类人” ：P

分享一下端木同学觉得很有趣的网站，all about statistics~

FlowingData
http://flowingdata.com/
a cool blog about data. run by a PhD candidate at UCLA

Wordle
http://www.wordle.net/
"Wordle is a toy for generating 'word clouds' from text that you provide."
The font size of the text is based upon the frequency of the word appearing in the provided document.

GapMinder
http://www.gapminder.org/
"Unveiling the beauty of statistics for a fact based world view."

TED -- Ideas worth spreading
http://www.ted.com/
"Riveting talks by remarkable people, free to the world"
In particular, check this out:
Hans Rosling shows the best stats you've ever seen
http://www.ted.com/talks/lang/en ... u_ve_ever_seen.html

Netflix Prize
http://www.netflixprize.com/
"The Netflix Prize seeks to substantially improve the accuracy of predictions about how much someone is going to love a movie based on their movie preferences. Improve it enough and you win one (or more) Prizes. Winning the Netflix Prize improves our ability to connect people to the movies they love. "

The BellKor Team from AT&T won ,000 Progress Prizes in 2007 and 2008 as part of the Netflix Prize competition, a M open competition to build the best recommendation algorithm using Netflix data.
http://www.research.att.com/~volinsky/netflix/

Part.II

六月十七日 周三 09:35PM 448，International House, Berkeley
宅在屋子里。Dr.V是从UW毕业的。
K同学很可爱的说。周一晚上跑去一个免费的Salsa课程，五六个人都有去。这辈子头一回跳舞，小纪念一下。K讲中文的时候超级可爱，不过除非明年都到伯克利来念书，再见面的概率接近于零。
I think I'm becoming more and more GEEK.
我对跑去SF玩一点兴趣都没有 - -b R的吸引力反而更大些。
今天是第四天。8:45AM -- 5:00PM，减去中间一个小时左右的午饭时间，这样的七个小时也是很繁重的脑力劳动。和之前做的dose response modeling刚好相反，现在不是数据不够，而是太多了。想要找出变量间的关系或是画出有意义的图，都要下一番功夫。
I love statistics. I love research. I will go for a Ph.D. Deal.

六月二十日 周六 05:42PM 448，International House, Berkeley
四点钟workshop画上了句号。和Ken，Boris在校园里溜达了一圈，其他的同学们都奔SF玩去了。We are outliers. Took a couple of pictures. I save the Sather Gate for next fall :) Wish my dream come true~
You make the decision, you pay the price. The best sentence to describe me yesterday is "there is too much blood in my caffeine stream." I was as quiet as usual in dinner. "Sheldon Cooper". Smart and sweet. I think it's okay to go wild once a while. One year.

The motto of UW is "LUX SIT". The motto of UC Berkeley is "LET THERE BE LIGHT".
Fiat lux is the motto of and also appears on the seals of thirteen educational institutions, which includes both UW and Berkeley.

/**

The Latin phrase fiat lux, from the Latin Vulgate Bible, is typically translated as "let there be light" when relating to Genesis 1:3 (Hebrew: "יְהִי אוֹר"). The full phrase is "dixitque Deus fiat lux et facta est lux" ("And said God let there be light, and there was light"), from the Greek "και είπε ο Θεός γενηθήτω φως και εγένετο φως" (or kai eipen ho Theos genēthētō phōs kai egeneto phōs), from the Hebrew "וַיֹּאמֶר אֱלֹהִים, יְהִי אוֹר; וַיְהִי אוֹר" (or vayo'mer 'Elohiym yehiy 'or vayehiy 'or).

Since fiat lux would be literally translated as "let light be made" (fiat is from fieri, the passive form of the verb facere, "to make" or "to do"), an alternative Latinization of the original Greek and Hebrew, lux sit ("light - let it exist" or "let light exist") has been used occasionally, such as in the motto of the University of Washington.

**/ from Wikipeida "Let there be light"

Part. I.

六月十三日 上午八点三十二 星巴克，西雅图机场
想申请的事情已经想了很久了。并不是简单的分数权衡和奖学金的问题，而是我这辈子，或更准确些，在我余下的人生里，究竟想做些什么。“人生的意义是什么？”很多年前就开始在想的问题，很抱歉，到现在我也说不清楚。至少了解了我不喜欢怎样的生活，也算是一点点进步了。
讲到自信，在人生转折的时刻，自信水平超过我的正常状态。报考XMU，转学UW，这次的summer program，尽管都有很多的不确定因素，但我相信我会如愿的。是因为前期调查详细，是因为对自己实力的客观认识，还是因为两者的有机结合？样本数太小，没法做95% hypothesis test : P
这次summer program是否可以看作是pre-selected？大多数人应该是junior升senior的时刻，下半年就要申请。在四月初到年底递交申请表的这段日子里，各方面变得更好的概率远远大于变差的概率。有stat grad program的学校一般都会设置undergrad major，这次应该都有通知到stat majors。每所学校只选一个人，24所学校想必都是很好的学校，至少stat program相当强。这样的stat major在他/她的学校应该是第一，综合考虑。在这个时刻看来，他们是竞争对手；但长远看来，他们是统计界的未来之星。认真做事，不论结果如何。当然，现实会复杂得多，the level of significance of my analysis is far greater than .05.
需要想明白的问题：how much do I love statistics? Do I really want to get a Ph.D degree?
尽管有些许不如意不顺心，我并不后悔三年前的决定。不然我也不可能进入UW学习，现在的我也不可能去伯克利参加summer program。那么明天呢？Do you want to be "Dr. Zhu"?
小猪加油～

下午两点三十一 International House, Berkeley
顺利抵达，顺利入住。居然跟我去年来这边住的那个hostel的风格是一样的，当然条件要好很多。单间，公用洗手间和淋浴，要用钥匙开门才行，我还没找到淋浴间在哪里 - -b 有发饭卡，午饭小尝试了一下。饭后小溜达了一圈，一间living room样子的大房间很古典，当时有位人在弹钢琴。顺着街道走下去，宛若回到了一年之前。
This is so unreal. I mean it's too great to be true. I went to UW one year before I applied. I thought there might be some hint from Life. Now I'm at Berkeley. Will I come here again next fall? Transferring to UW is not that hard after all. But I have no idea about my chance of getting into Berkeley. When I was applying for this summer program, I decided that if I got in, I would apply for Berkeley. Now if I cannot get in 2010, I will try again for 2011. This is where I want to be.
在Student Store转了很久，不晓得该买些什么礼物。毕竟不在这个学校念书，穿有Cal logo的衣服会很奇怪。看见一件Tee上写着“BEAT STANFORD” ：P 回房间在走廊里看到三个女生和一个男生正准备向外走，看起来很年轻超级有活力。我果然是老了。除了不能上网之外，我现在所做的事情和在家没有什么区别，想起TBBT他们在北极探险基地的生活了，让我们宅下去。

晃到phds.org弄了一份自己的排名，可以对不同的因素加权的，例如我很看重placement rate, faculty quality, female and minority rate, etc.
The result is as follows:

Rank	Department	Your Score	Yrs (-2)	Job (5)	Ed Eff (5)	Cite / Fac (3)	Fac Qual (4)	Support (5)	Grnt / Fac (5)	Fem (3)	Min (3)	For (3)	Stu / Fac (-4)
Adjust your weights →			-+	-+	-+	-+	-+	-+	-+	-+	-+	-+	-+
1.	Statistics, University of Washington-Seattle Campus	32.3	4.9	95%	3.9	57.1	4.0	98%*	1.6	47%	0%	74%	1.5
2.	Statistics, Columbia University in the City of New York	31.6	4.1	100%	3.4	48.9	3.4	100%	1.2	29%	6%	94%	1.2
2.	Statistics, Harvard University	31.6	4.7	85%	3.8	61.1	4.2	99%*	2.0	43%	7%	64%	2.1
4.	Statistics, University of California-Berkeley	30.3	5.2	90%	4.3	19.3	4.8	97%*	2.3	40%	0%	66%	1.6
5.	Statistics, Cornell University	24.2	5.3	83%	4.1	17.3	4.4	100%	1.5	46%	0%	62%	0.8
6.	Statistics, Stanford University	23.5	5.0	85%	4.4	53.3	4.8	100%	1.7	31%	0%	63%	3.1
7.	Statistics, University of Minnesota-Twin Cities	21.1	5.8	100%	3.7	12.6	3.9	99%*	1.8	43%	4%	43%	1.6
8.	Statistics, University of Wisconsin-Madison	20.1	4.9	82%	4.1	30.4	4.1	100%	1.5	35%	0%	86%	2.1
9.	Statistics, Carnegie Mellon University	19.1	4.7	92%	3.8	19.9	3.8	100%	2.1	29%	0%	59%	1.9
10.	Statistics, Purdue University-Main Campus	17.3	4.9	88%	3.6	9.1	4.0	100%	1.8	26%	0%	79%	1.5
11.	Statistics, Oklahoma State University-Main Campus	16.5	5.4*	92%*	3.2*	16.6*	3.3*	98%*	1.1*	25%	25%	50%	1.9*
12.	Statistics, Northwestern University	15.3	5.4*	92%*	2.9	42.9	3.1	98%*	1.9	14%	0%	71%	0.6
13.	Statistics, Yale University	14.4	5.7	99%*	3.4	16.2	3.6	100%	3.0	38%	0%	69%	4.0
14.	Statistiscs, University of California-Los Angeles	14.3	4.7	73%	3.8	56.9	3.9	98%*	1.4	29%	0%	57%	1.4
15.	Statistics, University of Chicago	14.1	4.9	94%	4.1	9.6	4.3	99%*	0.7	39%	0%	70%	2.3
16.	Statistics and Biostatistics, Ohio State University-Main Campus	10.7	5.7	100%	3.2	5.9	3.2	100%	1.9	62%	0%	50%	2.6
17.	Statistics, University of Rochester	10.2	5.3	100%	3.2	17.5	3.1	100%	0.7	43%	0%	71%	1.5
18.	Statistics, George Washington University	9.0	5.3	75%	2.8	9.4	2.9	98%*	2.6	40%	10%	50%	1.3
19.	Statistics, University of Missouri-Columbia	8.3	4.7	88%	3.2*	16.6*	3.3*	100%	1.1*	44%	0%	81%	1.9*
20.	Statistics, SUNY at Albany	7.5	5.4*	92%*	3.2*	16.6*	3.3*	98%*	1.1*	67%	0%	50%	1.9*

哈哈，按我的标准来评，第一是UW噢 ^^
不过我应该是没有机会在UW读phd的，除非我先去别的学校读个master。。。

09SPRING就这样结束了。总感觉自己还没有完成什么事情，时间就不见了。说我眼大肚子小也对。

修了四门有成绩的课程，已经知道了两门的成绩。372应该问题不大，虽然近一年多，我总是screw up finals - -||| ；担心342。M老大很有佛心，三个月就懒散地写了篇文理不通的review paper，他都不计较。
一开始还很怕S教授，但如A同学所说，跟他接触久了就会发现他人很好的。今天去拿作业，小聊了一会儿。原以为他还会继续教421，但换人了。425明年春季不会开课，因为系里预算的问题，不过这次也只有九个人上，感觉满遗憾的，也很庆幸自己死乞白赖地要上这门课 ：P 当时可是拜托了M教授帮我讲话的，开学前还收到S教授一封特正式的邮件，问我对R了解多少，并说考虑到班上其他同学对R已经很了解了，不可能占用上课时间来介绍R。一个星期自学速成，后面就是边写作业边琢磨了，也就这么过来。S教授教的很细致，当然前提是要适应他的教书风格  最后一堂课他送给每人一份纪念品，是他和别人在1987年写的paper，很有质感，呵呵。聊到Grad school，被S教授讲得对PhD有些动心了。

昨天和G教授见面。Java让我很头大，完全搞不懂GroupLayout是怎么回事，唉，不是所有东西都可以自学的。拿了Resume和成绩单，不晓得到时候课程会怎样算。问了一下会不会确保admission，因为觉得如果百分百了，我也就不用费劲儿去申请别的学校了。例如Tulane的BioStat算上退休和Affiliated也只有九个人，瀑布汗啊～ 答曰不要funding就问题不大。其实吧，不要钱，也就一年的事情，加之in-state tuition不贵，生活费又省一大笔，也可以接受。加之别的学校，就算我感兴趣，能拿到奖的概率也很小。
俗一点，只是看US NEWS的排名：

Math Specialty Rankings: Statistics

Ranked in 2008

1	Stanford University Stanford, CA
	University of California--Berkeley Berkeley, CA
3	University of North Carolina--Chapel Hill Chapel Hill, NC
4	Harvard University Cambridge, MA
5	Iowa State University Ames, IA
6	Duke University Durham, NC
	University of Chicago Chicago, IL
	University of Washington Seattle, WA
9	University of Florida Gainesville, FL

不想去the middle of nowhere，否掉ISU；不喜欢太冷的地方，对芝加哥不感兴趣；担心Florida的飓风把我吹跑。。。
唉，我毛病还真多。乖乖准备GRE，等考分出来了再做白日梦吧。

Stanford PHD program requires GRE math sub and asks for two sets of official transcripts. So does UNC. Duke only has a PhD program. I'm so screwed up by the transcripts : (

踏踏实实做事情吧，猪头～