Quick access to articles on this page:
more on the next page...
In my quest to better support to learn, I've again stumbled into a complicated, badly explained and unclear paper from my prof about discreet and time-memoryless channels.
Although it might be just me, but when I don't understand something from one source I like to diversify, and papers from Polytechnique (in french) are always a good snack :
http://www.enseignement.polytechnique.fr/profs/informatique/Nicolas.Sendrier/TI/cours6.pdf
Learning about Shannon's theorem in class I got curious and googled the guy.
One extract I found interesting in his wikipedia biography :
Shannon and his wife Betty also used to go on weekends to Las Vegas with M.I.T. mathematician Ed Thorp, and made very successful forays in blackjack using game theory type methods co-developed with fellow Bell Labs associate, physicist John L. Kelly Jr. based on principles of information theory. They made a fortune, as detailed in the book Fortune's Formula by William Poundstone and corroborated by the writings of Elwyn Berlekamp, Kelly's research assistant in 1960 and 1962. Shannon and Thorp also applied the same theory, later known as the Kelly criterion, to the stock market with even better results. Claude Shannon's card count techniques were explained in Bringing Down the House, the best-selling book published in 2003 about the MIT Blackjack Team by Ben Mezrich. In 2008, the book was adapted into a drama film titled 21.
Apart for inventing most of cryptography concepts, and doing chess IA, he also made a fortune from gambling and playing with stocks. Interesting.
My portfolio/vitrine/online resume... call it what you want, which is available on davidwong.fr, is now available on david.wong.emi.u-bordeaux1.fr as well. I thought that was pretty cool to have a bordeaux1.fr url. I think I can also have a univ-lyon1.fr since my account is still active but I can't be bothered looking at where it is.
Anyway, just this small piece of news in the ocean of bitcoin/litecoin news I've been posting here. Exams are coming soon and I should blog more about them than cryptocoins but yeah...
Done!
posted November 2013
So after a long night staying up and coding I finally handed in my project including my report in LaTeX.
I'm not really proud of what I did, I felt like I could have done much better if given more time (okay I slacked and I had enough time).
BUT, as I already said earlier, I've accomplished a lot and even though I'm done with this project I still kinda want to keep working on it.
Things that I've learned doing this class :
- C is awful. But now I know the basics. I wish we had one more project to code in C to really get it though.
- Makefile? Headers? I still don't really get the structure of a C project (and I'm ashamed).
- I know Linux! Okay I don't know Linux that much, but I'm getting really causy there. I installed debian on a VM and I'm considering setting up a dual boot on my laptop now.
- Emacs emacs! I was postponing learning it because I was afraid, and just forced myself to use it for this project and goshh am I fast when I use it. When I go back to Sublime Text I just want to C-M-F, C-A, C-K, C-Y...
- LaTeX! As a Math major I've always been ashamed not knowing it. Now that I got a taste of it I'm wondering if I should use it to write my book on.
- Svn and Git. I'm not a stranger anymore! And I use them for all my websites as well now :)
I think that's it, but I feel like I've learned a lot and I wished this course was a year thing rather than a semester thing.
The course is not over yet though and next week we'll dive into java for... a quick swim since it will be our last week.
I feel like I've been doing a hackaton these past few days trying to finish my sudoku solver. I had to hand it in 2 hours ago but still haven't finished... I really hope this won't affect my grade too much.
I've been learning a lot of Emacs, C, using gcov, gprof, LaTeX... I'm so confused right now and my code has became so dense that it's hard for me to debug it.
Yesterday, suddenly, I found something really stupid in my sudoku grid generation that I couldn't fix. A day after, I found the solution, randomly, fixing it created a huge load of other issues. I have been re-inspecting my whole code all day long and I'm stressed by this deadline that I already passed.
Gosh that is a hard course.
And... because of this, I missed a day writing on my new application. I was on a 9-day strike :(
I'm studying automata, it's sort of a "logical" subject that reminds me of studying mathematics. It looks cool, it only asks your brain to think, not to memorize, and you don't really know what's the real use of it.
If you want to take a peak at what I'm studying, you can find a similar course on Coursera given by Jeff Ullman from Stanford (yes, obviously I should have moved to the US and attend Stanford).
Well, someone nicely asked what I was thinking on Stackoverflow, and someone else nicely answered.
Réseaux
posted November 2013
I have an exam of Réseaux (Network) tomorrow and the slides of my prof are... how could I say this... not really clear. We have practical applications classes but they were... organized in the worst possible way. The subject did seem interesting at first but I felt like I learned nothing. Hopefully for the past few weeks I've been using the wonderful online course An Introduction to Computer Networks given by Nick McKeown and Philip Levis both very competent profs from Stanford. It seems like I should have gone there for my master of Cryptography :) Anyway, I'm doing with what I have here and I feel blessed studying Cryptography right when free online courses started becoming a thing.
The course is available here.
My Programmmation class first part is about coding a sudoku solver. We have to do everything in english, we have to commit with svn, we have to write a final report with LaTeX.
Every week we're given some vague guidelines and we have to dive deep into C to first, understand what we have to do, and secondly, find solutions in a language we've never really played with before. We have to turn in what we did every week, if our code doesn't compile it's a zero, if it does compile it goes through a multitude of tests that quickly decrease your grade (out of 20). Let's just say I spent many nights and early mornings coding and I started the first week with a 2/20.
It felt like a crash course, it felt unfair at times, but holy cow did I learn some C in a really short amount of time. Props to my professor for that, and I wish I had more courses like that. I might not get the best grade out of this course but I sure learn the most things there.
I've also committed everything I've done on a public git repo so everyone can see how it looks like here :
https://github.com/mimoo/sudoku
You can compile with make, learn how to use with ./sudoku -h
It can read sudokus of different sizes from 1x1 to 64x64 as long as it is presented like this :
#this is a comment
5 3 _ _ 7 _ _ _ _
6 _ _ 1 9 5 _ _ _
_ 9 8 _ _ _ _ 6 _
8 _ _ _ 6 _ _ _ 3
4 _ _ 8 _ 3 _ _ 1
7 _ _ _ 2 _ _ _ 6
_ 6 _ _ _ _ 2 8 _
_ _ _ 4 1 9 _ _ 5
_ _ _ _ 8 _ _ 7 9
