stevetjoa

Hooray, IPython 2.0.0 now has `Audio` in IPython.display. As someone who uses IPython notebooks for audio signal processing, this is huge. There were previous user-contributed ways to get around it, but now it's built in.

See http://nbviewer.ipython.org/github/ipython/ipython/blob/2.x/... for an example.

Good news: a moment ago, I was just able to install IPython 2.0.0 with `conda update ipython`. I'm excited -- it looks great.

Interestingly, if I then do `conda update anaconda`, it unlinks IPython 2.0.0 and re-links 1.1.0 after confirmation (which I decline, of course).

I was a reviewer for a Springer journal (MMIR). I contacted Springer in October 2012 to let them know that they store plaintext passwords and share the plaintext passwords with their users (e.g. "Hey, it's review season again. Please login to accept or decline your review assignments. By the way, if you don't remember your password, here it is! abc12345"). Over a brief email exchange, I carefully, politely, and concisely laid out the problems with doing so. All I got was a hilariously dismissive reply.

In October 2013, I received another email from Springer with my password in plaintext, and they CCed this email to another person!!

I declined my review assignments, asked them to permanently delete my account, and I haven't heard from them since.

As an EE, I used Matlab a lot for signal processing (image and audio). I had Matlab for Windows at the time. In 2007, I decided to slowly migrate my primary desktop environment over to Ubuntu, but I could not easily find (i.e., pirate) Matlab for Linux. Yes, I was aware of Octave and SciLab, but they are not enormously popular, particularly for niche fields that are comprehensively addressed by the Matlab toolboxes.

That's when I discovered Python (NumPy, SciPy, Matplotlib, IPython). As for the actual learning process, these three things helped me the most:

1. http://www.pythonchallenge.com/

2. http://projecteuler.net/

3. the emergence of Stack Overflow. I asked and answered a bunch of questions there.

I wrote more about my reasoning here in an old post (2010): http://stevetjoa.com/305/

Same here.

I've seen advertisements around SF for https://relayrides.com/.

How pleasantly surprised I am to find that the work of Amir Ali Ahmadi, a friend and former student at Maryland where I was his EE TA, appeared on Hacker News. We both took the same graduate-level class on optimal control; despite being an undergrad, he was one of the best students in the class.

The result is intriguing and reasonably accessible: for all (multivariate) polynomials whose degrees are even and at least four, determining convexity (of any type! strong, strict, regular, pseudo-, quasi-) is strongly NP-hard. To prove this, the authors equate the problem of determining convexity to the problem of determining the nonnegativity of biquadratic forms -- a known NP-hard problem.

Find the paper here: http://aaa.lids.mit.edu/publications. (Edit: direct PDF link http://mit.edu/~a_a_a/Public/Publications/convexity_nphard.p...) See Table I on page 18 for a concise summary.

I just bought the Das Keyboard Ultimate, and I like it a lot. The keys are firm and clicky, but not as firm as I remember the IBM Model M being. I'm slowly getting used to the unmarked keys, particularly the number row. The keyboard as a whole is weighted and doesn't move while typing. It's great.

For laptops, I love the Thinkpad. I returned an Asus Eee PC for the sole reason that the keyboard was unusable, and instead bought the Thinkpad.

Despite owning a MacBook Pro, I don't like the Mac keyboard and probably never will.

In case you haven't seen the Stack Overflow post on keyboards: http://stackoverflow.com/questions/687/keyboard-for-programm...

200 upvotes in nine hours?! Is that a Stack Overflow record?

Also won with white. AI let me fork its two rooks with my queen.

Careful; it is not yet clear whether this person is affiliated with LulzSec. You should add "alleged" or "suspected" to the title.

Quadratic growth is f(n) = n^2. Exponential growth is f(n) = e^n, or 2^n, or 10^n, for example.

One company, Remedium Technologies, is developing something similar -- a foam spray to control bleeding: http://www.remediumtechnologies.com

For those interested in similar websites:

http://isitsnowinginsfyet.com

http://diddukewin.com

As you read:

1. Write notes in a separate, clean notebook.

2. Ask questions. Write down the questions.

3. Read fewer books, but read them well.

For me, writing and asking questions significantly enhance reading comprehension.

Kinda sorta not really. Compressed sensing and sparse coding show that, under certain sparsity assumptions, you can perfectly reconstruct your original data with fewer bits than previously thought. It is a coding principle.

Perceptual hashes (or hashes in general) are used for fast indexing and retrieval. You cannot recreate the original data from a hash, pretty much by definition.

So hashes and coding algorithms both provide smaller representations of data. But hashes are used for indexing and do not provide the original data, or even an approximation thereof, while compressed sensing can.

If you're interested in this article, then you may interested in locality sensitive hashing (LSH), a randomized hash that has been used seemingly everywhere. I recently used it to speed up music source separation (papers pending).

The idea is similar to the one mentioned in this article, but more general. Unlike a cryptographically secure hash where x != y implies that h(x) != h(y) (collisions aside), LSH says that if x and y are "near", then P(h(x) = h(y)) is "high". This quality is important when doing robust similarity search. For example, if your image is noisy or rotated or scaled, you hope that you can still find the clean version in a database.

LSH has been used in many application domains including images, video, music, text, bioinformatics, and more. LSH is not directly comparable to a feature extraction algorithm such as SIFT.

[Edited for clarity.]

I do research in music information retrieval. See the ISMIR 2003 paper below. In short, it searches for landmarks in the spectrogram, hashes those landmarks, then compares those hashes against database hashes for temporal continuity. http://www.ee.columbia.edu/~dpwe/papers/Wang03-shazam.pdf

A seminal paper on audio fingerprinting is the one by Haitsma and Kalker. http://ismir2002.ismir.net/proceedings/02-fp04-2.pdf

RSS feeds. Most journals now have one. Examples:

JMLR: http://jmlr.csail.mit.edu/jmlr.xml

IEEE TPAMI: http://ieeexplore.ieee.org/rss/TOC34.XML

IEEE TASLP: http://ieeexplore.ieee.org/rss/TOC10376.XML

You could follow material posted online. For example, last semester, I sat in on this class: http://cbcb.umd.edu/confcour/CMSC828H-syllabus.html

If you are a bioengineering MS and know Python, then by studying the materials and journal papers, you may be able to do the assignments. At least it will give you an idea of the important problems and existing approaches.

Encouraging, but read this too, particularly the edit at the bottom, along with the link to ualberta.ca dated May 12, 2010: http://www.tweaktown.com/news/19694/has_cancer_been_cured_bi...

I study music information retrieval (MIR). Here are some of my bookmarks:

MIR Tools: http://www.music-ir.org/evaluation/tools.html

MIR How-Tos: http://www.ifs.tuwien.ac.at/mir/howtos.html

MIR Data Sets: http://grh.mur.at/sites/default/files/mir_datasets_0.html

MIR Toolbox (Matlab): https://www.jyu.fi/hum/laitokset/musiikki/en/research/coe/ma...

Million Song Dataset: http://labrosa.ee.columbia.edu/millionsong/

The "music" tag on Stack Overflow has some basic discussions on MIR: http://stackoverflow.com/questions/tagged/music

This excerpt is fantastic -- refusing to let Nixon's impeachment derail his path toward writing a novel.

When I find myself about to lose focus, I ask myself, "Ten years from now, are you going to remember that crappy TV show you watched for one hour on a Tuesday, or are you going to remember the moment your advisor kicked you out of his group?"

Investing. And related to that, risk taking.

"I watched two videos and didn't see anyone innocent getting killed."

What are you implying?

"Of course if I saw someone burying an IED, I'm not sure I'd be too sensitive either."

This article describes several unprovoked attacks by US infantry, the soldiers' total disregard for human life, and the investigation afterward which often led to inadequate punishment of the guilty parties.

Careful, planned murder of an unarmed 15-year old farmer by grenade and gunfire.

Severed human fingers as trophies.

Superiors who conveniently ignored this behavior.

The guilty soldiers' downfall after they beat a fellow soldier for snitching, then threatened to kill him, because "it's too easy."

Yes, JNMR is peer reviewed. http://www.tandf.co.uk/journals/nnmr

For those interested in more publications in my dear field of music information retrieval (MIR), see the ISMIR conferences: http://www.ismir.net/. All ISMIR conference papers are available for free! Great place to learn about the state of the art in music transcription, source separation, classification, search and retrieval, and more.

Also, Computer Music Journal (CMJ): http://www.computermusicjournal.org/

Yes, you're right. To clarify, I meant positions like copy editors.

I did not mean to generalize. I'm sure you're correct, particularly in other disciplines that are not as financially endowed. Economics 101, I guess.

In theory, perhaps.

In practice, this sort of already exists. And the process can take years. For example, some IEEE journals take six months per review round! (I know; I've just been through one!) Why does it take so long? I have no idea; reviewers just wait until the last minute, anyway.

Another problem is the definition of an "important" paper. That, itself, sparks debate.

One more problem is corruption. When reviews are not blind, bias can sneak into the review process. It happens. And it can distort all notions of what "important" papers are.