Updating with recent Udacity, Coursera and Edx courses, tutorials on Git... #36
Conversation
…Git, introductory links to Web APIs such as Twitter and web crawling, data journalism
|
This is awesome. Huge thanks for contributing! Questions (asked of all PRs)
Because Data Science is such a nebulous and undefinable discipline, maintaining strict standards for the core curriculum ensures its value as a complete curricular resource. However, your PR tipped the scale and catalyzed the creation of a few other very useful documents. Comments
General notes:
Thank you again! It's people like you who take the time to contribute that make this a truly valuable resource. It's new to feel such a responsibility to a living, breathing thing which is neither animal nor vegetable, but a convergence of many sharp and considerate minds. Collaboration is at the heart of (almost) all things true and good. I'd love to cordially invite you to a coffee in the Mission. Cheers! |
|
Hi Clare, sorry about the delay replying - I am travelling in Taiwan at the moment - and also that this message is in a comment, that doesn't feel quite the right place to put it. Yes I am taking many of these courses - see http://uk.linkedin.com/in/butlermh/ and please send me an invite. The reason for my submission is several people on the first course on the Coursera Data Specialization, Data Scientist's Toolbox, were discussing this resource and using it. This first course has been quite disappointing from my point of view, especially compared to the previous course "Data Analysis" that it replaced, but I guess there are a lot of people who want to study data science right at the beginning (for example they don't know how to use a shell, or git, what I would class as basic computer literacy) and this course is aimed at them (although it doesn't really go into these in a lot of depth, and there are much better intros if people need those type of skills). The problem is several of the MOOC providers are starting to commercialize, which to be honest is dropping the academic standards which is a shame because many courses used to be equivalent to the courses actually offered by the University. Now to specific questions: If I had to pick one I would recommend Data Analysis and Statistical Inference over intro to Stats purely because the latter is offered by Princeton so you don't get a certificate. But the latter is a very good course though (apologies to Andrew Conway here who gave a wonderful course and was a masterclass in good lecturing technique) I am just starting the Roughgarden Algorithms course so I don't know how they compare but from what I have seen the Sedgewick course is much more practical with the focus on implementation with pretty tight bounds on efficiency. I actually like doing more than one course, and hearing the information more than one way. One of my psychologist friends one told me the funny thing about memory is the more you know the easier it is to remember. The Data Journalism came up because I actually failed the Udacity Intro to Data Science course because they said the report I produced wasn't what they wanted and they wanted something based on Data Journalism. So for some people this is important ... I don't use Matlab myself, I work with Octave. Often the courses have people who only want to use one tool e.g. Python or R. It's great to be able to use a few. So I wouldn't use this in itself as a reason for not recommending a course (e.g. Andrew Ng's course uses Octave) like the Core Concepts in Data Analysis. It's only just starting and it's not one of my favourites .. but I note others have very different views to me .. so I guess in my submission I was trying to avoid being subjective. Yes sure there are many NoSQL databases ... but not so many courses. However I haven't done this course so I don't know the quality ... however I have seen even seasoned pros could do with a course / book on NoSQL because it means thinking in quite a different way ... Anyway I am afraid I have got to go ... Maybe more later .. best wishes! |
|
Sorry meant to type a message and actually just sent. I appreciate this Thank you, Sincerely On Monday, May 12, 2014, Ryan Jolicoeur rjolicoeur82@gmail.com wrote:
|
|
Hi Ryan, |
|
Hi Mark, Thank you for that information, I will certainly have to check that out On Monday, May 12, 2014, Mark H. Butler notifications@github.com wrote:
|
|
OK, finally get back to address the rest: I don't know about the Udacity Exploratory Data Analysis course, I haven't done it. I would hope it should be OK as it's done by people from the Facebook Data Science team (but the Udacity "Intro to Data Science" course was done by someone from industry too, but that was rather disappointing). One of my study friends has done it, he said it was OK but the assignment was a little frustrating. The three machine learning courses just came out, so neither my friend or I have done them. Also regarding $ most of these courses have a free version, it's just you don't get a certificate. You might want to add an extra dollar sign for Udacity as the fees tend to be double Coursera and you have to pay on a monthly basis. They do this because they employ tutors to help you (which is good for some people) but also makes them seem expensive for people who just want a certificate but might not need that level of support - I found having to discuss my "study goals" with my tutor before starting the course a bit frustrating really - he was a nice guy but it just didn't seem necessary. But I am lapsing into subjective comment ... Anyway it sounds like you have applied some editorial judgement, that is fine I guess if you think that is appropriate. |
|
Hi Ryan |
|
Hi Mark, Thank you for adding me on LinkedIn. I have several questions that I would On Monday, May 12, 2014, Mark H. Butler notifications@github.com wrote:
|
3 participants
Adding links to Coursera and Udacity Data Science specializations, recent EdX and Coursera courses, Data Journalism and Web scraping, and some other good introductory Python resources.
I have worked with many of these resources.