|
Hello. I purchased "Complete Machine Learning & Data Science Bootcamp 2022" on Udemy for cheap when it was on sale. I have no illusions that this will make me an epic hacker hired by amazon but I'm wondering if its worth a heck to do for my own interest? This is also the thread to discuss other online bootcamps etc, and whether they are worth a heck
|
#
?
Dec 31, 2021 00:15
|
|
|
|
| # ? May 3, 2024 17:24 |
|
|
FreeCodeCamp got me started about 5 years ago now - the front-end curriculum was good enough that I felt confident in building basic stuff and starting to experiment on my own. I think FCC did more for more me than my 1st year software courses. Codecademy was also useful... more to help solidify syntax when I was still struggling to remember basic structures. Probably better paired with another, more structured resource (like a cheap udemy course) to reinforce.
|
|
#
?
Dec 31, 2021 01:59
|
|
|
rodomontade posted:FreeCodeCamp got me started about 5 years ago now - the front-end curriculum was good enough that I felt confident in building basic stuff and starting to experiment on my own. I think FCC did more for more me than my 1st year software courses. are you a real software developer now if you dont min d me asking OP
|
|
#
?
Dec 31, 2021 02:31
|
|
|
I tried Codecademy for a bit before I ended up trying out a few other books and eventually going back to college to learn programming in a more structured environment. I can't speak for every single product out there, but the thing I didn't like about Codecademy (or really anything that has you working on an in-browser environment rather than your own computer) is that it teaches you the syntax and logic, but not how to actually run the poo poo locally - which is what I was interested in. Of course, now I know I was silly and it's not at all a huge ordeal, but I do think anything that wants to teach you to write code in whatever language should show you how to actually do something with it, including how to do it on your own machine rather than an in-browser editor. e: rodomontade posted:Codecademy was also useful... more to help solidify syntax when I was still struggling to remember basic structures. Probably better paired with another, more structured resource (like a cheap udemy course) to reinforce. this is the critical thing, I think - much better to compound multiple resources together. For example, I looked up that udemy course and it uses Python, but doesn't seem to explicitly teach it. This is the problem I've seen a lot of my students/friends run into - you're gonna hit something that looks like it should work, but it won't run, and having some knowledge of the Python language as its own concept rather than as "a thing to do data science with" will be extremely helpful. death cob for cutie fucked around with this message at 21:56 on Dec 31, 2021 |
|
#
?
Dec 31, 2021 21:05
|
|
|
Epsilon Plus posted:I tried Codecademy for a bit before I ended up trying out a few other books and eventually going back to college to learn programming in a more structured environment. I can't speak for every single product out there, but the thing I didn't like about Codecademy (or really anything that has you working on an in-browser environment rather than your own computer) is that it teaches you the syntax and logic, but not how to actually run the poo poo locally - which is what I was interested in. Of course, now I know I was silly and it's not at all a huge ordeal, but I do think anything that wants to teach you to write code in whatever language should show you how to actually do something with it, including how to do it on your own machine rather than an in-browser editor. It has a thing in it that is ostensibly teaching my python. But yeah it is having me learning it in repl.it for now but supposedly when you get to the real part it teaches all the poo poo to run it locally or whatever.
|
|
#
?
Dec 31, 2021 23:58
|
|
|
The browser thing is really tough, because in the real world or at a job, you're going to spend a lot of time messing with things that aren't the language itself. For python: organizing/structuring a module, dealing with versioning between them and their dependencies, crap like that. For a compiled language like C++ it's arguably worse because in my experience, picking and learning to use an editor, learning to use make or cmake, and compiling a trivial program is a huge hurdle for beginners that won't be captured if you do a browser course. Personally if I were going to go that route, I would start there to grasp the basic syntax, then move to working locally so I can really get a handle on how you use the langauge ecosystem. Would this be feasible or is it critical to use their system to get feedback on your answers? Alternatively, once you're comfortable with the basics, come up with or pick a project to do and start working on it locally. This is the best approach for me, but it's different for everyone! I did the Andrew Ng ML course on Coursera, and I found that format to be good - lecture videos, interactive pop-up questions, and assignments you had to complete on a local setup (MATLAB or octave) and submit to their automated grading system. It felt a lot like taking a true university course rather than web tutorial.
|
|
#
?
Jan 6, 2022 23:07
|
|
|
I'm almost done with the in browser portion of the data science course. I will keep you apprised...
|
|
#
?
Jan 10, 2022 17:18
|
|
|
styls trill epic posted:are you a real software developer now if you dont min d me asking OP Sorry late reply. Yeah, I've been working as a professional developer for a bit over a year plus another year of internships. I went to school though after a year of self-study* so I didn't pursue the self-directed route all the way to a professional level. I think it's possible to get to a pro-level without school, but it's gonna be a slog and I know personally I got better a lot faster once I could bounce ideas off of my peers. If you just want to a decent hobbyist level or as a supplementary skill for another type of office job, I reckon you absolutely can do it with courses like these. *(basically took the plunge since I figured the next provincial government where I live would ratfuck our generous student loan/grants, which they did)
|
|
#
?
Jan 11, 2022 18:38
|
|
|
I finished edX’s Harvard CS50x in December and found it pretty rigorous, although you do need to do a fair amount of outside reading to complete the exercises, as the lectures and shorts (intentionally) only give you about 80% of the information needed. Course covered fundamental design patterns and algorithms all using C, then Python, SQL and various web-specific bits and bobs. Caveat that I did it to learn how to do various specific things and not to induce anyone else to hire me, so while it met my goals someone else’s may be different.
|
|
#
?
Jan 14, 2022 02:45
|
|
|
I switched over to code academy being my primary learning method and its been ok so far.... I eel im learning at leaset
|
|
#
?
Jan 24, 2022 16:39
|
|
|
I'm doing codecademy, paid version (i got a big discount), but some of the courses are piss poor. The ruby course regularly tells you to print an operation's value to the console, but when you do it considers it a mistake. Looking at the actual solution just shows the original code that was already there with no modification. You have to ignore the instructions to check whether you did it right by printing to the console, then remove your print command so the stupid thing can consider your answer "correct" and then lets you move on. The site's okay for if you want to learn HTML and CSS and stuff, i dunno what the java and python parts are like There's a few combined courses like "full stack engineer" that combine bits and pieces of other courses on the site to create a single track, but it feels chaotic and like you're not really working towards anything concrete, it feels like doing flash cards
|
|
#
?
Jan 26, 2022 13:41
|
|
|
Another option I kind of forgot about was exercism, where a person will actually look over your answers to the exercises and give you feedback: https://exercism.org/. I did a little bit of a few languages, Lua I think? It was 2 or 3 years ago so I don't clearly remember how it went but I remember it being decent. Has anyone else tried this?
|
|
#
?
Jan 27, 2022 00:47
|
|
|
Shibawanko posted:There's a few combined courses like "full stack engineer" that combine bits and pieces of other courses on the site to create a single track, but it feels chaotic and like you're not really working towards anything concrete, it feels like doing flash cards this is a decent analogy for a problem I've seen my students or friends who're learning to code have: you know how to do what's asked of you, or how to complete the examples you're given, but not how to use that information in any other context. the best solution I have for that is, essentially, to try and start doing stuff independently as soon as is feasible. if you're following a tutorial that ends up with you writing a program to do X, see if you can make it do both X and Y with the same data, or see if you can figure out how to write that data to a file on disk rather than just spitting it out to the console. the sooner you start loving around and playing with your new tools, the sooner you'll become competent enough to create your own projects rather than doing what others suggest. styls trill epic posted:.... I eel im learning at leaset this is the good poo poo, if you feel like you're making progress you probably are. always jazzed to see someone learning and finding success in code, even if it's just for their own interest.
|
|
#
?
Jan 27, 2022 01:00
|
|
|
some more ruby today, just look at this:quote:Nil: a Formal Introduction then you get an array called creatures and you're supposed to enter a nonexistent value to see the nil result, so i put puts creatures["bullshitvalue"] which just results in an error. what you're actually supposed to do is just enter: creatures["bullshitvalue"] to access the nonexistent value. but this results in a blank console, there is no feedback or visible "nil" result like the instructions suggest, but it's the only way to proceed, it feels really weird because you don't actually get to see the nil result. after loving around for 10 mins figuring out what i'm doing wrong i went to the forums you see loads of people having the same problem because the exercises are all phrased like this, with no empathy for the expectations of the user and very little in the way of useful hints. html/css wasn't this bad though, it's just the ruby course that makes me look at the solution every time because nothing i do lets me proceed
|
|
#
?
Jan 27, 2022 02:59
|
|
|
 codecademy staff: "it's not our fault that our paid service sucks"
|
|
#
?
Feb 2, 2022 03:30
|
|
|
Any recommendations for a course that's structured on teaching Python completely from scratch but with data science in mind? I've followed some basic Python vids and want to start an actual interactive course but I want one that has all the examples, problem sets and projects be all setup towards using Python for data analysis and other stuff (bonus if it's in Jupyter Notebooks too). It seems most Python data science courses are setup with you already having done some introductory python courses which makes sense so you can learn how to declare functions and if then loops, but I'd prefer to do a course that has this all built in so it teaches all those things and then segues into the data science as soon as practical. I dunno if this is unrealistic but it's hard for me to keep my attention and focus if I cant feel like it's all connected with that ultimate goal in mind.
|
|
#
?
May 21, 2022 01:52
|
|
|
If you want to do data science you need to be comfortable with learning things that won't pay off for a while.
|
|
#
?
May 21, 2022 03:02
|
|
|
I taught myself purely with online resources and personal projects, and a few months ago I started getting paid actual dollars for freelance gigs, mostly through Upwork. This was more of a hobby for me to see if I could, rather than a serious, focused pursuit. So if I can do it anyone can. FWIW I actually really liked Angela Yu's iOS w/ Swift and Full Stack Web Development courses, would definitely recommend them as far as Udemy courses go.
|
|
#
?
May 21, 2022 10:32
|
|
|
ultrafilter posted:If you want to do data science you need to be comfortable with learning things that won't pay off for a while. Hey appreciate the sentiment but I'm not looking for career or professional development tips, just looking for courses that are the best fit for me and since I'm fine with paying I find having that structure from the beginning helpful.
|
|
#
?
May 21, 2022 10:55
|
|
|
Oysters Autobio posted:Hey appreciate the sentiment but I'm not looking for career or professional development tips, just looking for courses that are the best fit for me and since I'm fine with paying I find having that structure from the beginning helpful. I hope you're fine with a reply a month overdue, assuming the question is still valid. I can't recommend any specific courses, but Datacamp is probably my best since data science/analysis is their selling point and they have a number of Python-oriented courses with different structures; I'm sure one of them is at least closer to what you want. Other than that, I've found over time that Towards Data Science was more often than not the best Google result for specific problems I had in R/Python/ML. Some example articles that seem like they might be a fit for you include these two. Alternatively, if you have to learn Pandas specifically, you can try checking out R for Data Science for inspiration, as R and Pandas operate similarly in many aspects, though this is probably more of a stretch.
|
|
#
?
Jun 14, 2022 17:42
|
|
|
King Cocoa Butter posted:Another option I kind of forgot about was exercism, where a person will actually look over your answers to the exercises and give you feedback: https://exercism.org/. I did a little bit of a few languages, Lua I think? It was 2 or 3 years ago so I don't clearly remember how it went but I remember it being decent. Has anyone else tried this? yo i've recently been using exercism to learn elixir and i like it way more than codecademy compared to codecademy you have to do a bit more work to find the right answers yourself, so when you learn about elixir's list module for instance, you get an exercise which has something to do with lists and includes a basic explanation of lists, but it's up to you to find an appropriate solution to the problem by just reading the docs. i kind of like this because it doesn't allow you to simply regurgitate a version of an example answer and forces you to to learn actively but might be harder for if you're learning your first language. what i really like though is that it gives you some automated feedback on your solution, so that even if you find a solution to a problem that works, you're encouraged to make it more efficient/less hacky the interface seems to be a lot better than codecademy's, much more stable and it's easier to compare your solution to other people's. the people who "mentor" your answers also seem to be a lot more helpful than on CA. it's also free
|
|
#
?
Aug 30, 2022 21:50
|
|
|
Freakie posted:I hope you're fine with a reply a month overdue, assuming the question is still valid. Hey thanks, much appreciated. I have been leaning towards DataCamp so this helps. On another question, anyone got good tips on how to find good tutorials / books made in Jupyter notebooks? I used the gallery and I'll google dork it with the filetype but curious if there's any others. Oh and anyone got good resources for learning resources with lots of emphasis on heavy visuals? Especially for the more abstract stuff like data structures and algorithms. I like codephile and that one Math channel (3brownblue or something?) because holy poo poo does anything statistics or math make more sense to me with good visuals.
|
|
#
?
Oct 10, 2022 15:37
|
|
|
|
| # ? May 3, 2024 17:24 |
|
|
Oysters Autobio posted:Hey thanks, much appreciated. I have been leaning towards DataCamp so this helps. I wouldn't concentrate on tutorials specifically made in/for Jupyter Notebooks, since they're really just files that can run modular chunks of Python code in the same file. If you think you have a good resource, it should be pretty straightforward to adapt it to Jupyter. I'm also not sure I can give you a strong resource for visual learning, as much of the practice and intuition behind data science/analysis is so strongly tied to a mathematical foundation. 3blue1brown did help me a lot with the geometric/topological understanding of linear algebra, for example, but I'm not sure there's anything quite like that with statistics. Although his linear algebra series might be a good way to bridge the gap between math and data; maybe check it out if you have the time. If you haven't focused a lot on the foundation math and probability theory I would suggest checking out this website, specifically chapters 8 (and 1-5 depending on what you think might be relevant). The book does have some visual communication, but still requires you to understand the probability theory behind. I'm sorry I can't help more (and I hope this doesn't sound like gatekeeping) but it's difficult to really know what might be helpful if you're not taking traditional courses on these subjects, since a lot of these kinds of understanding have to be learned through exposure, practice and especially discipline if you're doing this as a self-study. Much of what I've said here isn't specific to Python, so it might also not be what you're asking for, but for me it's difficult to separate theory from practice when it comes to data science. You can also check out this book, for example, which at a glance seems like it has a lot of what you may want to learn (generally O'Reilly have decent or even good resources freely available through the authors' own sites).
|
|
#
?
Oct 16, 2022 14:02
|
|