dhammack

I'm not associated with the OP, but I have played with word2vec.

Word2vec can be viewed as dimensionality reduction. In some sense, the dimensionality of words is the same as the vocabulary size, if a one-hot encoding is used (which is most common). For word2vec specifically, the output vector dimension is a parameter specified during training, so you can choose it however you want. I suspect the OP chose something fairly large (>256) because of the performance boost.

Nearest neighbor methods work well with word2vec words: https://code.google.com/p/word2vec/#Word_clustering

The paper associated with word2vec is fairly easy to read and understand, even if you don't have a background in NLP or neural networks: http://arxiv.org/pdf/1301.3781.pdf

Hey, nice work! Can you explain the "comma delimited list" functionality any more? It seems (awesomely) similar to a hack I did a while back with Word2Vec which would pick out the word which didn't belong in a list.

My hack: https://github.com/dhammack/Word2VecExample

Kind of. LSI is a dimensionality-reduction method, which means it takes really sparse "bag of words" vectors and compresses them in a way which approximates the original structure. Word2vec gives each word a non sparse vector based on the hidden units of a neural network language model. The model is trained to predict the middle word from it's surroundings, which makes it learn how words are used in context.

There's an interesting python library [1] which implements AD as well as has some neat features like automatic compilation to optimized C. It's developed by the AI lab at the University of Montreal, and is pretty popular in deep learning circles. I've found it to be a huge time saver to not worry whether you screwed up your gradient calculations when doing exploratory research!

[1] http://deeplearning.net/software/theano/

Yes, I am. Here's what shows up in the dev console when I click "I'm an expert"

PUT http://www.interviewcake.com/api/v1/viewings/ 403

(FORBIDDEN) angular.min.js:99

(anonymous function) angular.min.js:99

n angular.min.js:95

l angular.min.js:94

DjangoRESTResource.(anonymous function) angular-django-rest-

resource.js:12

DjangoRESTResource.(anonymous function) angular-django-rest-

resource.js:18

$scope.updateViewing base.js:31

$scope.markAsComfortable base.js:13

$scope.feelExpertPress base.js:15

(anonymous function) angular.min.js:72

(anonymous function) angular.min.js:144

e.$eval angular.min.js:88

e.$apply angular.min.js:88

(anonymous function) angular.min.js:144

x.event.dispatch jquery.js:5095

v.handle

Just as a heads up, after I finish an interview question neither button works (I'm an expert or review later). Clicking on each doesn't seem to do anything. The only way for me to see additional questions is to head back to the homepage and re-click 'run some practice questions.'

Here's how to get MNIST, from the sklearn docs:

>>> from sklearn.datasets import fetch_mldata

>>> mnist = fetch_mldata('MNIST original', data_home=custom_data_home)

I think the handwritten digits dataset used in the presentation is just a subset of MNIST; MNIST is 28x28 and the handwritten digits are 8x8.

This is just an ordinary use of supervised learning; they just used Mechanical Turk instead of undergrads. The point of the human classification was to generate the labeled data for model training.

The talk they gave, "Deep learning for NLP (without magic)" was pretty good: http://techtalks.tv/events/312/573/

The vectors learned from word2vec are pretty amazing. A few days after the tool was released I wrote a script which uses the vector representations to figure out which word in a list isn't like the others [1]. Things like:

->math shopping reading science

I think shopping doesnt belong in this list!

->rain snow sleet sun

I think sun doesnt belong in this list!

etc.

[1] https://github.com/dhammack/Word2VecExample

I started using OneTab after seeing the post on HN, and I gotta admit that it has been really useful. Thanks for making it!

In my spare time, machine learning. I'm working through Bishop's book and testing out a few ideas of my own (a new way to ensemble models and a way for ANN's to communicate).

Awesome.

Is there a way to overlay the price? Also, it may be because it has just launched, but it would be cool to see a longer history with overlaid price.

Why stop at bitcoin? I'm sure the infrastructure you've developed can adapt to different keywords (company/product names).

New York.

Awesome, looking forward to interning this summer.

Looks like it keeps switching rules, but I've seen some rule 30 behavior in the animation. http://archetyp.al/other_pages/cae.js

Looks great! One suggestion (maybe you're working on this now) would be to have lessons for some of the characters which coding requires frequently (angle brackets, parenthesis, curly braces, brackets, etc.) For me at least, that's what slows me down when coding the most.

I had the same thoughts. In my experience, exponential growth in stock price is unsustainable, and usually results in a correction. Since the question is really how long do we have until a correction, I would say two to four weeks. TL;DR - Don't buy now.

I also agree, I do nothing illegal online so I fear nothing when it comes to Google "invading my privacy." I have a hard time understanding the people who obsess over leaving no trace online (unless they are doing something illegal). I also saw the story where Target used data mining techniques to predict pregnancy...all I can say is awesome.

A comparison of "conservative" versus "progressive" reveals some interesting things. Progressive leads to the Obama subreddit, where conservative leads to the Ron Paul and Rand Paul subreddits. The connections in each are ingriguing, for instance, the progressive has connections to environmental sub reddits, while the conservative has links to financial sub reddits. Interesting.

Someone will build the first quantum computer. I highly suspect that that person will also have to have a strong knowledge of quantum physics. Either way, it seems like a promising career field unless the theory for a quantum computer suddenly becomes impossible (which doesn't seem to be the case).

I appreciate the insight. Money isn't really the issue for me, and I don't plan to skip college altogether. I feel that with one year part time, and one year full time experience, when I do graduate I'll be much more marketable and my job prospects will be better. I also expect that I'll be more focused in college given that I've seen how what I'm learning applies in the real world. Either way - having more options can't hurt.

I have. Online is an interesting option - I was considering taking one or two night classes at UCF for subjects which are cumulative - like calculus.