| General > General Technical Chat |
| Can youtube STEM videos from universities replace higher learning? |
| << < (4/8) > >> |
| olkipukki:
--- Quote from: rstofer on October 11, 2021, 05:25:21 pm --- If you register and get a .edu email account, MATLAB is free! .... MATLAB (student version (FREE) or personal version (costs money)) --- End quote --- Must be US deal only. Never saw free MATLAB for students or home users (exclude who has access via college/uni), you either buy a book or pay around ~$40 for base plus ~$10 for each interested package. |
| daqq:
Concerning STEM universities: I do not believe so. IMO: Videos and courses can complement it - greatly. And can get you a lot of the knowledge you would get at a uni, possibly with a better explanation. But they lack: 1. Feedback - Testing is important. While the exact methods may be better or worse depending on the university/country, solving a complex problem that can't be fitted into A, B, C or D in a limited time frame with whatever resources you are given teaches you something. 2. Lab experience - Universities have some nice toys. Depending on your setup you may not be able to recreate it at home. And some of the experience is not so much of a technical nature, rather than a general organization nature - how to record measurement, organize data... 3. Talking and interacting with your peers - If you put enough smart people in a room, they eventually come up with something awesome. Or get fiendishly drunk or high in illegal herbage. Anyway, talking to people who are studying the same things that you are, are in the same situation, in the same context is valuable. 4. The boring stuff is important - Math was mentioned here as being problematic. Yes, most of the time you will not sure a lot of the higher math and can generally manage without a mathematical feel for it. But for a title and the advanced work that you may be expected to do, at least some of the boring stuff will come up occasionally. 5. Recognized title - If you get a Bc. or Ing. ("bakalar" and "inzinier" here) title, it should mean something. While it is true that there are "universities" that, let's say, are of dubious nature, they are quickly recognized and people who flaunt ther Ing. from certain schools will get ignored by any competent HR people. There are certain expectations that come with the title. True, it may be more diluted than it was in ye olden days, but it still means something. During my university years, there were times where I was this close to giving up on it. In retrospect, I was right to continue. Talking MATLAB: Try octave. |
| Kasper:
Sorry to hear about the Drs and meds. But it's good you've found a productive hobby. I think respect for 'youtube university' is growing and there are plenty of opportunities for people with good experience and knowledge even without formal credentials. Are you trying to decide if you should get a formal education or are you trying to decide if hobbyist learning is worth the effort? It sounds like you enjoy learning about STEM so that's good in any way you do it, even if it's just a low cost hobby that is good exercise for your brain. If you decide later that you want a formal education this could be a good headstart so you can enjoy your formal schooling and learn a lot while your classmates stress, cram and forget. If you have prior experience and maybe some projects to show off, that can help your resume stand out against your competition (classmates) when you all enter co-op or graduate. Can you get a technical job without formal schooling: yes. Will you get paid less? I dunno, maybe. But what about starting your own company, either to sell products or sell yourself (as a consultant)? Formal education might not be as important there. If you start a company, is it going to rely on government funding or investors? Some of them like to see formal credentials but some might not care. If it's hard to go to formal schooling, is it going to be hard to be an employee? You might find more flexibilty while working for yourself. I work for a tech startup and when we consider hiring design firms or consultants, we talk about what they've designed in the past and who they've worked for not what school they went to or even if they went to school. Having STEM knowledge is valuble in many ways. Having formal university credentials is also valuble. Nothing wrong with starting as a hobbyist and maybe getting formal education later. It also depends what area you want to work in. If you want to make consumer grade apps, some online courses could be plenty, if you want to design bridges or other life dependant grade things then formal education is probably required. If I were to start over, I think I'd be better off if I thought of something I wanted to make and sell and then learned whatever is needed to accomplish that. |
| Rick Law:
--- Quote from: Kerlin on October 10, 2021, 02:45:00 am ---I watched the 23 lectures from MIT by Alan Guth on inflationary cosmology. I believe he is the one who mostly conceived and proved the idea. I did not follow some of the maths as he some times used his knowledge of cosmology to introduce changes into the calculations and introduce new formulas. I decided to just follow as best I could as I didn't think he needed to prove it because other well know academics have already checked it. With the outcomes the LHC is providing and the advancement of current physics it seems that we are entering a new age of understanding. Details of these advances will be in the media. Albeit further down the news pages than the political, crime, football and baby competition winners. ... ... --- End quote --- (Just as I was previewing and editing before post, another reply by daqq brought up the importance of interaction. What goes here will serve as a "supplement" to daqq's point #3...) This actually reminds me about an additional draw back of "remote learning" which applies whether you are self-learn or "remote" on-line student due to whatever reason. That additional draw back is interaction. When you are at a lecture or in a university, surely you will have some interaction with others. Particularly useful would be Q&A from others to the teacher. Some questions may be same as yours, other questions may be something you just have not yet thought of. Besides Q&A, there is the "hall way discussions" or "things said by around the water cooler" where informal brief discussions occur. From that, you learn additional useful knowledge. Don't get me wrong, Physicists like Alan Guth and Lawrence Krauss are Physicist I much admired and still do. "Cosmic Inflation" is just a mean to interpret the information we are able to grasp. Neither space nor time are well defined in physics. Some would even argue that we don't even understand them. Consequently, space (which we can't define well) expanded in a very short time (which we can't well define) is a concept that we can't truly define well. But, that it is "a way to interpret the data" often got morphed into "that was it." In Q&A and informal discussions, that it is but "a mean to interpret the data" would surface. The same applies to LHC's "accomplishments". At the time of LHC's creation, many Physicist were ask, and repeated since: "If the Higgs particle is the only new particle that LHC could find, what would you think?" The answer was almost universally "that would be quite a disaster." You can see some of those Physicist on videos saying it on "World Science Fair" and similar shows. There was expectation that SUSY particles will come out; the "Standard Model" of Particle Physics would be verified. Not a single SUSY was found. Not a single one other than Higgs boson which was proposed in the 1960's! Many Physicist consider the last half century a time waster. In business terms: no new strategy, just plenty of new tactics. The generation of physicist like Eisenstein, Paul Dirac, Erwin Schrödinger, Max Planck, Werner Heisenberg, so on. They created new paradigm. They created new concept. Those concepts lead to new ways of thinking about Physics and new understanding. Since then, most Physicist merely developed new techniques. String Theory was to be the new paradigm, that didn't work (or if you prefer, hasn't work yet). The "Standard Model" of Particle Physics remains a mass of kluges... I suspect a few years of interaction (4 for college alone and perhaps more), one would have heard and know of these kinds of view points (or facts if you prefer). Lacking interactions with others and taking just a few on-line courses, many such important facts or view points would be missing. EDIT: "...said by the water cooler" replaced with "said around the water cooler" and added Werner Heisenberg to the list. There are more such Physicist for sure, but Heisenberg's Uncertainty Principal is too important to miss. |
| rstofer:
--- Quote from: olkipukki on October 11, 2021, 07:19:36 pm --- --- Quote from: rstofer on October 11, 2021, 05:25:21 pm --- If you register and get a .edu email account, MATLAB is free! .... MATLAB (student version (FREE) or personal version (costs money)) --- End quote --- Must be US deal only. Never saw free MATLAB for students or home users (exclude who has access via college/uni), you either buy a book or pay around ~$40 for base plus ~$10 for each interested package. --- End quote --- My 'personal' edition cost $149 plus $49 for every add-on package and I have several. But that's just the first cost because there is an optional upgrade fee of about 50% that comes up every year. I have about 8 add-on packages so annual 'renewal' is a bit pricey. Nevertheless, I really like MATLAB. These days, I'm playing with machine learning and MATLAB->Simulink->Deep Learning toolbox is a very nice setup. If I had a video card with CUDA units, I could add on the parallel programming toolbox and speed up the production by a LOT. All of this machine learning stuff can be done using any tool that handles matrices and vectors even, or maybe especially, Fortran. I haven't seen any literature talking about using Octave for this application but it is clearly capable of handling the math. The math isn't hard but the array size is often fairly large. Even digit recognition turns out to have a 784 element input vector (an unwrapped 28x28 array). Not the kind of thing you want to do by hand! |
| Navigation |
| Message Index |
| Next page |
| Previous page |