cttet

You can do both.. But it is hard.. https://www.youtube.com/watch?v=zt0OQb1DBko

Don't have the chance to watch the video but highly disagree with the title.

Math as a language is just a DSL which carries a lot of historical burden that is not friendly at all. Programming languages can achieve same level of abstraction with more clarity. Math instead added too much specific syntactic sugar(bad is multiply(b, a, d), log a is log(10, a) or log(e, a)), operator overload (/ for divide and dy/dx for differentiation operator, same grammar for scalar and matrix multiplication, one of which is communicative while the other is not), and complicated design (integral would be much clearer if we write integrate(target=ax, over=x, from=0, to=10) instead of the original one).

Yes. Even when Mongolia occupies China, in textbooks it is called the dynasty of Yuan. But it is not invented by the communist party, it is told in this way following the tradition by the Chinese scholars. https://en.wikipedia.org/wiki/Scholar-official

Yeah, that kind of thinking still exist in China. Even in popular fictions the ancient masters are always better than younger generations. https://en.wikipedia.org/wiki/Wuxia

I agree. China is more isolated geographically and hence less diverse than Europe and hence it is still quite different. It is definitely not the only culture that passed that long though.

I am just sharing thought base on personal experience.

I never mention 5000 years of History, only 1500. I am trying to not related with any politics here.

By your reasoning if there exists some propaganda then my personal experience should not be trusted?

In addiction, the long history is not created by CCP but following the Chinese scholar tradition. It is correct that CCP used it for creating nationalism, but they did not fabricate the whole thing.

To know more about Chinese culture not from [politic fueled](https://en.wikipedia.org/wiki/John_Ross_(academic)) books, I suggest you to read The [Problem of China](http://www.freeclassicebooks.com/Bertrand%20Russell/The%20Pr...) by Bertrand Russell. It is authored by a well-known scholar and written before the CCP exists in case you accuse me of suggesting you propaganda. The opinions by the book may be too old but at least the facts are unbiased IMO.

For the first part, I still think that Chinese culture is quite special with culture that passed on for that long.

As a normal Chinese with standard Chinese education till 10th grade (then move to a foreign country), I can understand fragments of writings carved in stones 1500 years ago in museums, which I felt quite amazing.

I am not sure what is your programming style in Pytorch. As people recommended and in most tutorials I see the sequential approach, where a small mistake of data preprocessing would lead to clueless errors in a completely irrelevant line.

YOLO is a quite standard feed-forward model in my opinion. I mean the math part, which I am more concerned with.

I have never used Theano before, my idea from it is that Tensorflow followed its static graph approach.

You don't often need to debug, especially if the model can be checked on compile time. Think static types for programming. For Tensorflow all the data types and tensor dimensions are checked before loading any data, it the math is derived correctly then it is not necessary to even debug.

If data and model are mixed, it often resort to line-by-line debugging to zone out the real problem, which often takes more time.

I just like Tensorflow better. For building new models, the graph is complex and errors are unavoidable. There is a separate compile time for Tensorflow and errors will be found before the data come in. Tried pytorch before, the error messages are usually not helpful at all and often leads to clueless debugging for hours.

For trying out deep learning, or build on existing models, pytorch or keras may be easier to grasp. But when making new models that involves a lot of math, the Theano/Tensorflow is more helpful IMO.

The main difference is that for search you need to specify an image to emulate it, while the RL method aim to generate new ones after observing some examples.

The figures that you have shown are sea otters, and they look quite different from other otter species.

Yeah the new talent tree make it less carry-ish.

Especially when I know that WhatsApp system is built in ErlangVM, which is famous for its ability to maintain being functional across updates..

I would change the font to sans-serif

The social structure shift is really not about preference, it is about collective power and violence. Although people may prefer the hunter-gatherer lifestyle, the neighbor farm-based civilization will conquer and enslave them.

It is also the same for technology/AI. They may not improve our quality of life, but they allow more powerful armies that would defeat those who resist them.

It's sad but it's true.

Maximum likelihood ¯\_(ツ)_/¯

Agree, I have learnt both Clojure and Haskell on codewars when I was bored in lab. Another advantages is that the typical problems in codewars often does not involve learning extra libraries or IO, which often distracts me from learning the language itself.

How does it compare to itchat?

Bayesian involves integration over priors, and conjugate prior/MCMC are tools for integration, which are implementation steps IMO, rather than essential part of understanding the Bayesian approach.

Power law indicates an exponential distribution, which is often taught right after normal distribution in most probability classes. Normal distribution usually considered as white noise, hence a fixed value + white noise is often Normal, which is quite common as well.

The lack of a proper package management system is the reason I stop using and recommending Matlab.

It was my favorite because of its excellent documentation and community support.

Is there any example of PGM applying to this field?

Poor Butters

"This is unacceptable because mathematics is written as English is written in complete, grammatical sentences." Unfortunately natural languages have many flaws and will easily falls in contradiction like Russell's paradox. Even formal mathematical axiomatic systems encountered Gödel's incompleteness..

Maybe something like constructive category theory would give us a better way to formally put ideas in ...

Music itself is about repetition, without repetition is just noise..

I am not an evolutionist or biologist to answer the first part, but when a theory(model) cannot explain some phenomenon does not imply external factor. It could also possible be caused by randomness.

But design is not only about looking good. It may look nice but with terrible UX.

I saw the video, he supports it.....

Yay, break the hard mode in 9 attempts! Kinda like Sudoku but more flexible.