Jos Sturm was born in Rotterdam, The Netherlands, on August 13, 1971.

He graduated in August 1993 in operations research from the Department of Econometrics at the University of Groningen, where he had also been a teaching assistant in statistics and a research assistant in econometrics. From Sept 1993 to Sept 1997, he was a PhD student at the Econometric Institute and the Tinbergen Institute of the Erasmus University in Rotterdam, under supervision of Shuzhong Zhang.

Jos spent the academic year 1997/1998 as a post-doctoral fellow at the Communications Research Laboratory (CRL) of McMaster University, Hamilton, Canada, as a TALENT stipend of the Netherlands Organization for Scientific Research (NWO). At McMaster University, he was a member of the ASPC group of Zhi-Quan (Tom) Luo.

After his post-doctoral fellowship at McMaster, he returned to the Netherlands, and from October 1998 to January 2001, he was a lecturer at Maastricht University, at the Department of Quantitative Economics, in the group of Antoon Kolen.

In January 2000, his PhD thesis was awarded the Gijs de Leve prize for the best thesis in operations research in the years 1997-1999 in The Netherlands.

In February 2001, Jos was appointed as an Associate Professor at Tilburg University, and a fellow of CentER (Center for economic research). In July 2001, he was awarded the prestigious Vernieuwingsimpuls (Innovation) grant of the Netherlands Organization for Scientific Research (NWO).

Jos was the editor of the newsletter SIAG/Optimization Views-and-News of the SIAM Activity Group on Optimization (SIAG/OPT), and a council member-at-large of the Mathematical Programming Society (MPS).

His scholarly works include more than 30 papers, and his PhD thesis was published in edited form in the volume `High performance optimization', Frenk et al. eds., Kluwer Academic Press, 2000. He was also the author of a widely used optimization software package called SeDuMi.

He is survived by his wife Changqing and daughter Stefanie, and his brother Pim and parents Frans and Els.

这学期3门课,3个中国老师.

Luo的Teaching Skill无疑是三门课中最NB的。

把一门Engineering为主的课程讲得Theory和Application相得益彰，实为难的。

Luo的口语也是significantly good。

由衷地赞一句，这位85年加入全国数学培训班，拥有MIT Applied Math PhD的Tom Luo，NB。

曾经很烦这个小城,最后却发现已经习惯了这里安静的生活.

看来我是一个适合隐居的人,呵呵

人到什么时候都不能为了爱情而丧失了尊严.

爱是平等的,建立在平等基础上的爱才会有前途.

当到了其中一方牺牲自己的尊严才能挽留的时候,爱情其实已经到了尽头.

祝福我的好友,希望她能找她的幸福.

刚刚才知道CVPR2008里有一篇和自己当初投ECML2008被拒的文章如出一辙。因为都是利用当
时渐渐冷下来的LLE的Local Linearity做文章，那Linear DiscriminatAnalysis的Model做
文章。而且我不但提出了CVPR里的方法，还提出了一个原理相通性能接近但是速度更快的方
法。但遗憾的是，ECML只有第一个审稿人认为Accept，其他都是Reject。

不禁感慨。
1，CVPR 2008的确好中，同样的创意我投ECML2008 rebuttal的机会都没有。当然ECML非盲
审，学霸们很容易瓜分文章，无疑对新手雪上加霜。
2，很多看似很牛的人很多情况其实就是会写而已，他们的方法不见得多新颖多有创意。同
样也能看出现如今学术圈的浮躁。审稿的趋势是用缺点排除文章，而不是看优点选取文章。
想想当初Pattern Recognition Letters的审稿人审我其他一篇文章的时候，态度真是不错
，尽管有文法和组织方面的缺点，还是着重点评了我的方法本身，并最终同意接受了我的文
章。
3，其实这个方法的几何解释存在问题，我收到拒稿信之后，就试图去解决这个问题，今天
读CVPR2008的文章，发现里面居然用了我当初也用的错误的几何解释。
4，CVPR2008的文章里根本没有提最近两年里大家在LLE的Model上进行的改进。我在ECML的
文章里详细描述了LLE最近几年里在降维问题里的应用和变种，最后ECML的审稿人认为我的
contribution很微小。
5，CVPR2008那篇文章纯属运气好或者人家关系牛，我ECML被拒也正常，自己有几把
、刷子心里有数，文章水不水自己挺清楚，就是不太平衡，呵呵。不过如果当初真中了这篇CVPR，我今年申请的情况可能会被彻底改变。

Dear X,
I am very disappointed to see that once again you have chosen to simply copy to solutions from the solution manual.  I have asked you not to do this several times and do not understand why this continues.   I have explained to you that the solution manual often skips steps and I want a key prepared for students that is explain clearly and completely.
Are you capable of doing the problems on your own? 
Could you please explain to me why this continues?
Sincerely,
Angela

没法出去的愤怒版本

Dear Angela,
Yes, while making sure I understand everything in it, I copied from the book directly, but even that took me over two hours. Please appreciate the work. I don't think I can write better solution than the book did and that already looks very detailed for me. Would you please give me some examples of problems where I need to further elaborate? Second, I don't go to your lectures, so I found I very often do the problems in a different way. Copying the solutions is a good way for me to know about this.
To answer your question, yes, I am able to solve the problems, but if I write it on my own, I am afraid it couldn't be any more detailed and well explained than the solution book.
In case you don't know, J and I are working at least 30% longer than all other TAs in the department. I am not gonna say anything about this as long as it is not going too far beyond the 20 hours and it hasn't taken too much time from my own study, because I love teaching. I understand you are an excellent instructor with very high standards, but to maintain that kind of standards, perhaps you need a third TA for the extra work. So far as I know, other TAs usually don't have to write the homework key or make the quiz or hold review sessions, while we are doing all these. Not to mention Calc 2 needs more time for preparation.
Please let me count the MINIMUM hours I spent on teaching every week:
Discussions: 6
Office hour: 3
EXAM Grading+review session: 7  (or Quiz+modification+key+grading: 6)
HW: 3
TA mtg: 1
Preparation for discussions: 3
TOTAL: 23(22) (again, minimum)
I want to offer extra time to teaching, but sorry I can't, 'cos I am also quite busy with my own course work this semester. It is understandable that we work over time occasionally, but not every week, please. The worse thing is that, after you offer more, they ask for more.

最终发出去的温柔版本

Hello Angela,
I copied from the book, but I did read through all the answers before I wrote them down, which looked very detailed for me. Especially for those word-describing answers, I don't think I can write any better than that. Also quite often I do the problems in a slightly different way, that would bother students a little bit.
I know you are more experienced in teaching and certainly know better what the students and graders need to understand the problem. Could you please tell me which problems need what kind of further explanation?
Because I want to give a detailed and perfect answer for the students, I used the book's way of illustrating the problems. I have done my best, although it's still not good enough. Would you help me fix that? To be honest, if I wrote it myself, it can only be worse. Sorry for not being able to meet up with your standards. I know that's all for students' good.

- Show quoted text -

这些日子除了焦躁还是焦躁，一想到会一直等到4月份甚至更晚就更加焦躁。

漫长的等待以U Toronto的拒信拉开帷幕，这个学校我很喜欢，因为Machine Learning Group十分强大。没想到这个常常拖到6月份才发拒信的学校居然早早的就把rejection用平信寄到了我家。

Univ. Georgia的Admission接踵而至,UGA是综合50名左右的学校，但是CS专业排名里是看不到的。这个学校的老师是早就联系好的，这个老师认识国内的老板，所以对我的背景和国内老板的推荐还是比较认可的。但是他还向确认一下我的编程能力，所以发了一个project让我做，这下可好，这个Project是使用Visulization Tools Kit来实现大脑皮层的主要沟壑的曲率估计。一个数据格式的说明文档就要19页。但愿能快点把这个做完吧。

今天早上刷到了Arizona State Univ.的Admission,第二个AD，ASU综合100名左右，CS专业排名50左右，算是不错的Backup。这个学校的老师也是很早就开始联系的，发信告知，对方说“Great. Thanks for the notes”。对于RA只字未提，不知道能有什么结果。

Purdue已经发了大批的Offer,Dartmouth开始面试这些牛校都没我的份。只希望2月份的下半程再多些AD，让我也有更多的机会在3月份拿好offer.

继续焦躁的等待

花了3个星期拿到了驾照，不容易不容易。终于不用蹭车了，生活也没有以前那么无聊了。

ICDM的审稿意见下来了，3个weak reject，这是今年拿到的第三个拒信。前面有CVPR08和ECML 08。一个挺好的idea，一折腾再折腾，到7月份ICML，开会，有人已经发表了一篇更好更全面的论文。我的idea一下子沦为半残，呵呵，我也不知道说什么好。国内硕士毕业，到美国来再读个硕士，反倒没有在国内的日子里有成果，好像混日子一样。

这么想进学术圈，图个啥呢？我也不知道，反正过去是很狂热，觉得自己实在搞创新，在作研究。现在看看，在哪里不是一回事？工作又何妨，人家也没比faculty差到哪里。就更不要提工作的硕士和在读的PhD，那待遇简直是天壤之别。

有人基础好，肯努力。有人天份好，有想法。有些人导师好，做事快。我呢，基础一般，天分马马虎虎，也没有人指导，何必非跟他们去挤独木桥呢？我曾经相当一名researcher,现在沦为一个发牢骚的怨妇。我也不知道该怪谁了，难道该怪我自己选错了路？

在深圳的时候，自己一个人打拚，以为到了美国找一个好一些的团队，和一群有思想的同龄人一起打拚。好好干四五年，然后拿个PhD学位。可现在发现，一切居然这么遥远。真是自作孽不可活啊，我选了一条不适合自己的路，或者说自己没有适合这种追求的运气。

眼看已是9月底，来美已经一年多了，仍然是没有人管，导师也不知道在忙什么，反正是忙不到我这里。混吧，混吧，在哪里不是一回事... ...

唉！

作理论比不过数学或者统计的，说是科学，其实就是人家玩剩下的。

做应用有显得没有什么技术含量，说是工程，实际也造福不了人类。

再回头来看看机器学习的鼻祖，泰斗，都是统计大家，似乎没有谁是计算机专业出来的。

当然也可能是因为计算机这个专业本身很年轻，培养的人才较少，但就目前计算机专业的课程设置和研究能力来看，真是看不到什么希望啊

rt

感谢人民感谢党

跑了4天多的程序结果没了... ...

3个审稿人的意见如下：

1，第一个认为文章新颖，可以接收。

2，第二个认为文章存在太多问题，列了一堆细枝末节的问题，“实验不足”，“说明不够”，这些模棱两可的话也都写了上来。应该是写明了要拒我。

3，第三个人为文章很有趣，但是有些问题都看不懂，给了个中性的评价。第三个人肯定是外行，因为我在backgroud里照抄了行内很有名的一个人的一段说明，他居然有若干个关键的部分没有看懂。

三个审稿人都指出，文章语法不够流畅，有很多错误。看来有必要对写作进行一下集中的加强。嘎嘎。

总体来讲虽然没达到最高目标，好，中，差各一个的评价对于这个顶级会议来说，并不算太坏。希望改一改有机会在ICDM上露一手。

今年上半年最后两篇文章，一定要在7.1日之前截稿。嘎嘎

这是02级本科的Ziyang Liu的Publication List

Ziyang Liu and Yi Chen: “Reasoning and Identifying Relevant Matches for XML Keyword Search.” In 34th International Conference on Very Large Data Bases (VLDB) / Journal of Data Management Research, Vol. 1, 2008 (to appear).

Yu Huang, Ziyang Liu and Yi Chen: “eXtract: Generating Query Biased Result Snippet for XML Search.” Demo Description. In 34th International Conference on Very Large Data Bases (VLDB) / Journal of Data Management Research, Vol. 1, 2008 (to appear).

Yu Huang, Ziyang Liu and Yi Chen: “Query Biased Snippet Generation in XML Search.” In Proceedings of 26th ACM SIGMOD International Conference on Management of Data, 2008 (to appear).

Ziyang Liu and Yi Chen. “Answering Keyword Queries on XML Using Materialized Views.” In Proceedings of The 24th International Conference on Data Engineering (ICDE), 2008

Ziyang Liu,  Jeffrey Walker, and Yi Chen: “ XSeek: A Semantic XML Search Engine Using Keywords .” Demo description. In Proceedings of 33rd International Conference on Very Large Data Bases (VLDB), 2007

Ziyang Liu and Yi Chen. “Identifying Meaningful Return Information for XML Keyword Search. ” In Proceedings of 26th ACM SIGMOD International Conference on Management of Data, 2007