Profile

Cover photo
Suresh Venkatasubramanian
Works at U. Utah
Attended Stanford
Lives in Salt Lake City
1,003,550 views
AboutPostsPhotosYouTubeReviews

Stream

 
My academic lineage traces back to someone who did NOT have a Ph.D. Talk about evidence for a real imposter syndrome :)
Link to Piled Higher and Deeper
2
2
Jeff Erickson's profile photoMoshe Vardi's profile photoElisabeth Grieger's profile photo
 
My great-great-grand-advisor was Bob Floyd, who didn't have a PhD (despite Raimund's repeated offers to introduce a cycle into the family tree).

But he had a Turing award, so no impostor anything.
Add a comment...
 
This "proponents of the latter" includes every one in computer science !!!
 
The PhD thesis: what format shall it be?

The classical 300 page doorstopper, now seen by some as somewhat bloated?  Just a string of disconnected publications?  Or the "integrated" approach, now seemingly en vogue, in which some publications by the PhD students are padded with a proper introduction, a literature review, and conclusions?

Proponents of the latter claim that much time is saved, and successful PhD candidates are better prepared for the rat race of postdoc positions, having already some publications under their belt. Critics point to the fact that students somewhat unlucky in their publication schedule will not graduate in time, and that the whole process pushes science further  towards projects easily publishable. And we have already enough of these. And how to judge multi-author papers in this context?
Should the foundations of a 21st-century academic career still be built on the traditional model?
5 comments on original post
5
Adam Smith's profile photo
 
Yes, indeed. Nothing says "synthesis and critical review" like a good staple. 
Add a comment...
 
the only parenting advice blog you'll ever need. 
 
What if you don’t have any original ideas but still need parenting tricks? Find out on today’s Mothershould:
On Mothershould, Grace Manning-Devlin helps viewers build a legal case when their parenting tricks are used by parental plagiarists.
1 comment on original post
1
Add a comment...
 
My contribution to the discussion. 
Over on Windows on Theory, there's a solid discussion going on about possible changes to the format for STOC 2017 to make it more of a 'theory festival'. As Michael Mitzenmacher exhorts, please do go and comment there: this is a great chance to influence the form of our major conferences, ...
1
Add a comment...
 
This book is worth getting on the basis of the introduction.
 
"This book complicates the idea that trolls, and trolls alone, are why we can’t have nice things online. Instead, it argues that trolls are born of and embedded within dominant institutions and tropes, which are every bit as damaging as the trolls’ most disruptive behaviors. Ultimately, then, this is why we can’t have nice things, and is the point to which the title gestures: the fact that online trolling is par for the mainstream cultural course."
A new book details the success of trolls in turning the media against itself
5 comments on original post
2
Add a comment...
 
I read the title and thought, "I wonder if +Mikael Vejdemo Johansson has seen this". Then I clicked through and realized it was his article :)
 
Sartorial mathematics and why we should care about the computational complexity of internet protocols
View original post
6
Kevin O'Bryant's profile photoDave Clarke's profile photoroux cody's profile photo
3 comments
 
+Dave Clarke This is not the point he's making. For all intents and purposes, the set of protocols that are accepted in, say, an ssh authentication are not Turing Complete. The implementation of the protocol has to recognize valid protocols, and the recognizer is usually written in a general purpose language, but that's not the issue.

In general, you want only a small subset of all possible sequences of queries/messages, e.g. "request signature" -> "receive signature" -> "send data" is allowed, but "request signature" -> "send data" -> "receive signature" is not. If your implementation does not quite conform to this intended subset, it may be hard to know you have a bug, since all normal users are only sending messages according to the allowed subset.

But some malicious user may recognize that your language allows for some bad sequences and exploit that. Excluding this behavior is, in a sense, what a lot of research in security protocol verification deals with. See, e.g. Paulson: http://www.cl.cam.ac.uk/~lp15/papers/Auth/jcs.pdf
Add a comment...

Communities

7 communities
 
Mathematicians have a 3% change of being automated, and professors have a 4% chance. Interestingly, while lawyers have a 3% chance of being automated, judges have a 40% change of being replaced by a robot. Predictive sentencing, anyone ? 
 
Will your job be around in the future? We take a peek at the research.
1 comment on original post
3
3
Arun Krishnakumar's profile photoHelger Lipmaa's profile photoSaravanan Thirumuruganathan's profile photoElisabeth Grieger's profile photo
 
"The researchers admit that these estimates are rough and likely to be wrong."
Add a comment...
 
Yes yes yes ! I HATE 'trapezoid' :)
 
US vs. UK: Mathematical Terminology, e.g. is that shape a trapezoid or a trapezium?  http://mathwithbaddrawings.com/2015/05/20/us-vs-uk-mathematical-terminology/
I’ve encountered some surprising disparities, each of which prompts the obvious question: Which way is better, the British or the American?
13 comments on original post
7
1
Alok Tiwari's profile photoHelger Lipmaa's profile photo
 
"Pythagorean theorem" doesn't make sense to me. (We don't have Gaussian theorem, Stokesian theorem, Eulerian formula, ... etc.)
Add a comment...
 
Why I like ITA...
And thought I'd answer it by mentioning an event that I really enjoy attending. I didn't post it as a comment because it's a little out of scope for the blog post itself: it doesn't make concrete recommendations so much as relay anecdotal evidence. The Information Theory and Applications ...
5
Add a comment...
 
 
Tycoon wants to have scientists sacked

At the University of Oklahoma. Lets set the stage, and introduce the players:

Harold Hamm, billionaire and owner of Oklahoma City-based Continental Resources, is what people call a "major donor" to the university. In 2011, he gave $20 million for a new diabetes center. (Named after him, of course.)

David Boren is president of the University of Oklahoma and also member of the board of directors of Continental Resources.

Austin Holland works at the Oklahoma Geological Survey (OGS), hosted at the university.

Larry Grillot is the dean of the college where the OGS is based.

And now the play:

Holland is called to a meeting with Boren and Hamm, in circumstances he describes as "a little bit intimidating".

During the meeting, Hamm expresses his concern about researchers publicly discussing certain research results.

Later, Hamm meets dean Grillot. After that meeting, Hamm expresses in an email that he would like to sit on the search committee for the next OGS director.

Afterwards, Grillot writes to colleagues that "Hamm is very upset" and that "he would like to see certain OGS staff dismissed".

Epilogue:  Grillot stands up for his staff members. Nobody got sacked.

So everything has ended well, in the short run, and you can guess how many further donations the university will get.


http://www.bloomberg.com/news/articles/2015-05-15/oil-tycoon-harold-hamm-wanted-scientists-dismissed-dean-s-e-mail-says

http://www.bloomberg.com/news/articles/2015-03-30/big-oil-pressured-scientists-over-fracking-wastewater-s-link-to-quakes
The billionaire CEO of Continental Resources told a dean at the University of Oklahoma that he wanted earthquake researchers dismissed
View original post
2
2
Parasaran Raman's profile photoRajarshi Guha's profile photoHelger Lipmaa's profile photo
 
He looks evil too.
Add a comment...
 
 
#NSF   #CISE  core solicitation deadlines for this fall:  Medium and Large Proposals in Sept., Small Proposals in Nov. NOTE CHANGES.
NSF 15-079. Dear Colleague Letter: Announcement of Changes to Submission Windows for the Computer and Information Science and Engineering (CISE) Directorate's Core Programs as well as the Secure and Trustworthy Cyberspace (SaTC) Program. May 13, 2015. Dear Colleagues: ...
View original post
2
Add a comment...
 
A long-ish rumination on the different ways of modeling higher order interactions. In brief, they're all of the form of "a graph is a special case of X" for different values of X. 
This year I'm one of the PC Chairs for SIAM Data Mining (along with Jieping Ye), and so I've been spending time in decidedly-not-sunny Vancouver. Being a PC Chair, or even being on a PC, is an exercise in constant deja vu: I hear a talk and think "Where have I heard this before" before realizing ...
2
Arnab Paul's profile photoJeff Erickson's profile photoChandra Chekuri's profile photoSuresh Venkatasubramanian's profile photo
5 comments
 
I disagree with the claim that spectral methods are not about representation. For problems like clustering for example, the only real question is: what's the representation ? The algorithms are then mostly an afterthought. For example, it doesn't make sense to use a spectral representation unless you believe at some level that a "manifold" view of the graph might be useful.

The issue with tree width is more interesting. In my view, the use of treewidth to "bound complexity" is purely a play for tractability - granted. But in a "parametrized sense" you are also making a choice (via representation) about which parameterization is important, and they're not all equivalent. 
Add a comment...
Communities
7 communities
Work
Occupation
Researcher
Employment
  • U. Utah
    Associate professor, present
Places
Map of the places this user has livedMap of the places this user has livedMap of the places this user has lived
Currently
Salt Lake City
Previously
Aarhus, DK - Stanford, CA - Philadelphia, PA - Morristown, NJ - New Delhi, India - Berkeley, CA
Story
Introduction
CS prof, interested in algorithms, geometry, data mining, clustering
Education
  • Stanford
  • IIT Kanpur
Basic Information
Gender
Male
Other names
geomblog
That this restaurant has an overall rating less than 4 is a travesty. Skip Finca's overrated food and preserve your ear drums: Cafe Madrid is a much more intimate (read: quiet and charming) Spanish fine dining experience, with possibly the best service I've ever had in Salt Lake City. Call ahead if you want the paella: it's worth it.
Public - 3 months ago
reviewed 3 months ago
3 reviews
Map
Map
Map