These are chat archives for FreeCodeCamp/DataScience
discussion on how we can use statistical methods to measure and improve the efficacy of http://freeCodeCamp.com
@jacobbogers @bitgrower I would say the same: DS is not only about d3.js, actually... visualization is part of it. What would you like to do instead?
...ultimately ... predictive analysis ... but, one thing I learned was how eye-candy can railroad ideas into folks' minds, even over better, less well presented ideas ... (granted, the eye-candy to 'less well presented' is a continuum) ... that's not to say I'm not un-interested in data viz ... just that there's more to it ... IOW -- I've seen people use (presumed?) data visualizations to mis-lead people (and that's outside of my comment about the power of just the eye-candy ...
One of the most powerful ways to mis-lead people is to give them "partial data" (eg, a comparison chart among various auto-responders) ... where relevant features might be left out ... or a competing product not mentioned at all ...
from @jacobbogers, re:book ... Multivariate Data Analysis
Joseph F. Hair Jr. et all ...
Awesome ... the one I just picked up via one of the used book seller sites (though this said 'new' -- and is the 7th edition) ... I was looking at a couple of other books to pick up as well -- but I couldn't find nearly the same good deals on them :) They're ones recommended by the Vancouver data blog guy ...
m-henderson sends brownie points to @samai-software :sparkles: :thumbsup: :sparkles:
@bitgrower Thanks for the feedback, captain! Totally agree. I wish we would have more data for prediction.
You should check the whole MIT series for sure about AI by P. Winston.
Other video we were recommending about ML is from Google series (python + tensorflow).
Good luck with multivariate! It is a very nice topic. I would suggest to use R for the exercises. I haven't see that book yet but I have a couple here already. If you are serious with that stuff I would join for a refresh.
evaristoc sends brownie points to @bitgrower :sparkles: :thumbsup: :sparkles:
@m-henderson in order to reduce the friction for creating some viz for the project we are also helping to get the data using R. We can either help you with the script or even provide you the data in json if required.
I would say that if you are learning R, you should try to get the data yourself, as an exercise, and ask for help here. :)
@koustuvsinha contact you later about the Radar-ish project? I am hesitating about it... Still going to a meetup tomorrow at nodeschool and I will check the pertinence of embarking on a project as the Radar-ish...
Otherwise we could concentrate on our previous projects instead... send you later a PM, today or tomorrow.
Free Code Camp's main advantages are that we're accessible to busy adults who want to change careers.
One of the most noticeable differences in this curriculum than other Code-Camp/School environments is the level of support: Using the chat-rooms for help, guidance, and assisting fellow coders gives you a better understanding of the code, easier access to explanations, and it's in real-time so you can question the confusion.
@m-henderson: @hjh17 is right when talking about contributions to FCC, although he is giving you just the big picture. ;) There are a lot of places where you can contribute at FCC! Just ask. I suggest to start with the wiki (ask @Rafase282) and then progressively get involved in more difficult tasks where you will need to do a bit more. When actually trying to push a commit you will have to work also at things like, for example:
@m-henderson by participating in more complex tasks you will get the experience of how teams work on the correct maintenance of development projects, which is part of the curriculum and a requirement for contributing in any existing collective project out there.
@m-henderson in our case (DSR, Survey 2016) is a bit simpler because we are not asking for commits at the moment, but only to use the dataset for analyses and visualizations.
The repo for the Survey 2016 project is there to check:
If you are interested in working on some viz:
The final projects will be featured in a website that it is currently under construction by someone at CodeNewbie. I think Quincy will take care of the content.
m-henderson sends brownie points to @evaristoc and @hjh17 and @samai-software :sparkles: :thumbsup: :sparkles:
evaristoc sends brownie points to @hjh17 and @m-henderson :sparkles: :thumbsup: :sparkles:
Thanks for the links, @evaristoc -- yeah, I already went thru the short Python/Tensor flow videos -- Josh Gordon's "promo" when Google was livestreaming Google I/O got me very excited about it ... (I hadn't really been all that excited previously)
...though I'm beginning to think I may need to brush up on my math ... (it's been a LONG time...)
@evaristoc, do we have an R room, then ? That's something I don't seem to have in my universe of educational resources (granted, I haven't pursued it, either...)
@SamAI-Software -- LOVE the animation ... what's your "recipe" (what languages are you using?) ... it's very cool
@all -- if you haven't looked at any of Hans Rosling's TED talks -- HIGHLY Recommend them ...
bitgrower sends brownie points to @evaristoc and @samai-software and @all :sparkles: :thumbsup: :sparkles:
:warning: could not find receiver for all
koustuvsinha sends brownie points to @bitgrower :sparkles: :thumbsup: :sparkles:
bitgrower sends brownie points to @evaristoc :sparkles: :thumbsup: :sparkles:
@bitgrower :) :) ... I would always prefer a real cookie instead...
People: the following is for those who are looking to work on topics related to Social Media analytics...
yeah ... if anyone was in doubt where big tech is going -- Google I/O's keynote pretty well demolished that notion ... :)
kinda like ... first we had desktops, then we had laptops, then we had mobile ... now we have chat as the "central" app ... it's not how much time people spend on FB ... it's how much time they spend in FB messenger ...
@bitgrower radicati is a company that has been analysing this niche market for long time, so they are probably using more data but keeping a small range for presentation purposes...
More about Instant Messaging market forecast...
Of course, @jacobbogers! I can see the purpose. Could you please share again the link here? I think it could benefit everyone who is interested. I would take the link into other groups too if I would be you... I am also interested in the content and the dynamics... Maybe we can translate some into this room so we can formalise study groups as part of the activities of the DataScience room?
This is for my own reference (a former project for a old Digest that I am planning to resume...):