wutf

The only way I would fund GIMP is if that funding went directly towards a better UI. It's not worth improving the algorithms given how hard it is to use.

I didn't why they got hung up, but I'm not a mathematician.

Given the information they have, it seems they can, starting with the 2D Julia set, use an evolutionary procedure to fold the object into a 3D shape (potentially not unique) such that each point on the Julia set has the correct MME. i.e., you try a random fold and have a particle follow many random walks, and measure the MME at each point. If the MMEs are closer to their true values, you keep the fold. After this is all done, apply a smoothing algorithm that minimizes the number of folds.

It's an IQ test.

"is very rich broad and compelling"

But does it support commas?

This isn't actually true. During construction of the test they throw out questions that don't categorize you. It has discriminative validity even when it looks close.

What you have to understand is that statistically the models are very similar (you can compress them both into one unified model that does what both of them do quite well). However, the ways the models are constructed makes them useful for different things. The Big 5 is primarily useful for academics, and the MBTI is primarily useful for the rest of us.

If you are a logical positivist and scientific realist you'll never be able to grok this. As a utilitarian I understand that science is the process of making something that does something you want done.

You're no statistician or psychologist. Those loadings are huge.

You can see the original mapping here: https://en.wikipedia.org/wiki/Myers%E2%80%93Briggs_Type_Indi...

PS: I'm the one who added that table to Wikipedia - many years ago.

They both have strengths and weaknesses. Strengths of the MBTI include the text descriptions, which are very valuable, and that it focuses on positive psychology. Of course, if you rotate the MBTI onto the Big 5, you see that it does in fact measure neuroticism, although not that strongly.

IIRC the first one was done in the original McCrae & Costa paper introducing the Big 5. Will dig it up in a bit. Edit: https://en.wikipedia.org/wiki/Myers%E2%80%93Briggs_Type_Indi...

The covariance of the Big 5 and MBTI is high. This means that if you give both inventories to the same population you can do a factor rotation of one onto the other with much of the variance being preserved. This has been demonstrated multiple times in the academic literature. I have never read an article against the MBTI that mentioned this, and that's because the people who write these articles do not understand statistics. The MBTI is approximately as valid as the Big 5.

Also live in Boulder, also avoid bike lanes. Absolutely terrifying.

Time to rethink RethinkDB.

Why are there two copies of this thread? https://news.ycombinator.com/item?id=12466983

You are just one person.. I don't have to do anything to access articles other than instantly visit the article page and download the PDF in one or two clicks.

also 2015

The book Where Mathematics Comes From demonstrates that the concept of infinity itself is finitistically reducible via conceptual metaphor. This really isn't that surprising though, is it? Our experience is grounded in the interaction of concrete, finite objects.

There are an infinite number of such concepts in "machine learning." Just because you liked this one doesn't make it a good interview question. If anything, you are failing at theory of mind - as evidenced here, most people aren't making lots of use of the concept. When interviewing, don't be a dick. Test people for general ability and willingness to learn, not stupid questions you got out of a random textbook.

Torrent the papers. Why hasn't this been done?

Just read it - quite good!

love it!

Thanks much!

The video is largely useless without the visualizations, given that the video is exclusively focused on visualization...

Learning doesn't need to be done in realtime... it can happen async.

Training is expensive, prediction is fast.

I love how I got super downvoted for an idea that is quite possibly central to the future of computing. It's as if HackerNews never realized that they themselves are machine-learning-based scheduling algorithms. Guess that one is going to hit them pretty hard.

Prediction can be extremely fast.

"I don't see how you can predict workloads off of historical data"

... Think about that part, then:)