Around the blogs in 80 hours: Compressive Sensing and More

After featuring Randy's blogging yesterday in Of Paradigm Shifts and Regime Changes, here are some other news of interest to Compressive sensing and beyond:

• Bob asks a question about a compressive sensing/l1 minimization classifier after the dubious results in his search for a good music genre classifier. He is right, you can get 92% classification capabilities but it ain't about Music genre: Music genre recognition results
• In a similar vein, Seth tells us about a not-so-optimal database for cancer research. I am sure current databases are a whole lot more accurate.
• Looking back at what happened last week and the sudden changes in terms of technology development, David Gorski's words (focused on preclinical research) take a specific importance

• :".... One of the most difficult aspects of science to convey to the general public about science-based medicine (and science in general) is just how messy it is. Scientists know that early reports in the peer-reviewed literature are by their very nature tentative and have a high probability of ultimately being found to be incorrect. Unfortunately, that is not science as it is imbibed by the public. Fed by too-trite tales of simple linear progressions from observation to theory to observation to better theory taught in school, as well as media portrayals of scientists as finding answers fast, most people seem to think that science is staid, predictable, and able to generate results virtually on demand. This sort of impression is fed even by shows that I kind of like for their ability to excite people about science, for instance CSI: Crime Scene Investigation and all of its offspring and many imitators. These shows portray beautiful people wearing beautiful pristine lab coats back lit in beautiful labs using perfectly styled multicolored Eppendorf tubes doing various assays and getting answers in minutes that normally take hours, days, or sometimes weeks. Often these assays are all done over a backing soundtrack consisting of classic rock or newer (but still relatively safe) “alternative” rock. And that’s just for applied science, in which no truly new ground is broken and no new discoveries are made...."

• Danny has a very nice interview of Michael Mahoney on random projections among other things. Michael  has been organizing the MMDS meetings for the past few years. It is a must attend meeting. Same goes for the GraphLab: Big Learning meeting as mentioned earlier.
• Dusty tells us how to build frames from codes
• Pierre let us know about FREAK, Fast Retina Keypoint, a competitor to SIFT. As soon as the code is there, I'll make a longer mention of it. The paper is here.
• Muthu tells us about the current exodus of researchers from Yahoo! If you know some on LinkedIn, you may have already seen that trend. 
• John talks about Both new: STOC workshops and NEML
• Laurent tells us about Hyperbolets.
• Alex ask "Are there perturbations that preserve incoherence and give nicely conditioned “submatrices”?"
• Vladimir points to the same topic Randy has mentioned earlier and which I wrote about yesterday. I realize that the CMOS/CCD business is competitive but I will never understand the level of acrimony of some of the comments. Ouch. 
• Greg mentions his course at MIT: Learn about Phased Array Radar Systems by making your own in 5 days, an MIT Short Course
• Anna defines Universal Hash Functions
• Cable And I are continuing our experimental adventures using Robust PCA to find UFOs on the Moon and Make Invisible that Not So Invisible Mercedes. We also wondered if this approach could be used to perform a better lucky imaging.
• I wondered if compressive sensing should not be directly applied with wavelet chaos as opposed to polynomial chaos.and featured an implementation of High Speed Compressed Sensing Reconstruction in Dynamic Parallel MRI. 
• I also updated the Just on the right side of Impossible (part 2) but also had to write about advanced Matrix Factorization in view of the press and blog entries on the subject.
• Finally, John talks about Random is as random does

Talking about competitive technology, I just saw this plan for a drill-robot for Europa. Why would this be a consideration when you can "simply" use RTGs to do the job. I am not sure I understand the constraint for not going that route.

Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.