41
votes
Meta-post for the "a layperson's introduction to..." series
What's this?
This post contains all entries of the "a layperson's introduction to" series. I will keep this thread up to date and sorted. This means this post is an excellent opportunity to try out the bookmarking feature!
Physics
Quantum Physics
Basics of quantum physics
| Topic | Date | Subtopics | Author |
|---|---|---|---|
| Spin and quantisation part 1 | 01 Nov 2018 | spin, quantisation | @wanda-seldon |
| Spin and quantisation part 2 | 03 Nov 2018 | superposition, observing, collapse | @wanda-seldon |
| The nature of light and matter part 1 | 16 Nov 2018 | light, matter, wave-particle duality, photoelectric effect, double-slit experiment | @wanda-seldon |
Material Science
| Topic | Date | Subtopics | Author |
|---|---|---|---|
| Spintronics | 18 Jul 2018 | spintronics, electronics, transistors | @wanda-seldon |
| Quantum Oscillations | 28 Oct 2018 | quantum oscillations | @wanda-seldon |
| LEDs | 10 Nov 2018 | leds, electronics, diodes, semiconductors | @wanda-seldon |
| Spintronics Memory | 22 Jun 2019 | spintronics | @wanda-seldon |
Classical physics
Thermodynamics
| Topic | Date | Subtopics | Author |
|---|---|---|---|
| Thermodynamics part 1 | 07 Nov 2018 | energy, work, heat, systems | @ducks |
| Thermodynamics part 2 | 13 Nov 2018 | equilibrium, phase changes, ideal gas | @ducks |
| Thermodynamics part 3 | 24 Nov 2018 | @ducks |
Computer Science
Artificial Intelligence
| Topic | Date | Subtopics | Author |
|---|---|---|---|
| Genetic Algorithms | 18 Jun 2019 | algorithm, genetic algorithm | Soptik |
Group theory (along with most post-highschool math topics) is pretty hard to do a layman's intro for since it's so driven by definitions and terms that need to be digested and worked with a bit.
I guess you could do an overview of some generally interesting bits, but you'd definitely have to skip any real explanations in favor of hand waving, making vague observations rather than anything rigorous.
I've had a few years of the stuff, so I might be able to help out if someone decides to write something up.
Hm, I don't agree with the second paragraph, it feels a bit hyperbolic to me. For example one could prove Lagrange's theorem without all the preliminary language, effectively defining the things without having to write "the proof of lemma 1.1.1. follows from definition 1.0.1 and theorem 1.0.0" a la Bourbaki. I'm not saying it's easy (or that I'll do a good job at it) but I think a good communicator and expert who decided to put some effort into it would be able to explain things well.
Now of course one thing mathematicians love to do is "oh but this theorem that we made in the context of orbits and symmetries is also applicable to this more abstract structure like a monoid or a matroid or whatever look how cool that is!" but honestly that is not adding rigor in my opinion. I enjoy it as much as any other mathematician but you can do a lot of stuff with Z/nZ, do it rigorously and just say "this generalizes to other kinds of groups but the proof doesn't and it becomes much harder", that doesn't make it less rigorous. A good book in this sense (although it doesn't necessarily agree with me) is "Proofs and Refutations" by Lakatos.
I plan to write about it, talking mostly about actions on a set (symmetries, to make it more easy to connect with physics), leading to Burnside Lemma (explaining in the middle some of the required concepts), and then finishing with a veeery low level intro to Sylow theory. How does that look to you?
For Burnside's Lemma, you can probably get across the general idea, but the main equation involves syntax that represents some fairly technical definitions.
|X/G| is the order of the quotient group. Quotient groups in general are a bit hard to grasp, but I suppose if you limit equivalence relations to an informal idea of rotations or symmetries (rather than coset and equivalence class formalisms), then it's not so bad. Xg is a bit abstract too.
I guess when you make your objects concrete, then abstract algebra does become fairly straightforward. However, you've also lost much of its power and purpose and it appears like a weird representation of fairly ordinary things rather than a profound way of describing and relating structures based on their inherent properties. If made concrete, the isomorphism theorems become ad hoc proofs based on peculiarities of the objects involved. Basically, if you prove things based on their representation rather than their structure, then you aren't really doing abstract algebra.
Yeah, you are right but it still is, I think, part of the history of abstract algebra so in that sense it's not incorrect. It's like if I wanted to write about probability theory, I could just say "well a probability is a normalized measure, and conditioning is doing a Radon-Nikodym derivative, and you just need a filtration of the sigma-algebra to do this and that"... which would be rigorous and true and definitely easier if everybody shared the language. But I could also start from the Laplace rule, space of events, independent events, etc. And I don't think it would be less useful (for a layperson).
As you see this is a thing I have opinions on, and I used to think more like you but maybe having been very involved in outreach has made me change my opinion. IDK, judging by the posts about spin, quantum oscillations, spintronics, etc. I think something at that level can be done (and then in the discussion go deeper if necessary, or possible). We'll see how the group theory post goes :-P
I think you could give a general idea of abstract algebra by taking two example objects that have different representations and showing how they are actually isomorphic and how you can prove things about them simultaneously by leveraging theory proven for their common abstracted structure.
I didn't originally plan on adding enthalpy and gibbs energy since it's a bit more technical. I can write a bit about it, but I think that will be an addendum to part 3 (if all goes to plan). I'll try to remember to tag your name when that day comes.
EDIT: In general I don't mind going a bit more technical if people want it. That goes for all the stuff posted.
Just a general remark: I would recommend using letters for months in the table (e.g. 08-Nov-2018 rather than 08-11-2018) so that the date format is unambiguous.
I'd be interested in a layperson's introduction to audio engineering, something expanding on this post by @FreeLunch would be really neat
If this is going to be a Tildes-wide phenomenon, crossing multiple academic disciplines and areas of knowlege, what about putting it in ~tildes?
You don't need to summon him; I've moved this topic for you.
I ain't no discount-nobody! I'm a full-price Algernon! Hmph!
It would be great if someone could make post about neural networks, how they work and how to make one. But I'm not sure if it isn't too complex one.
I could make post about Genetic Algorithms (I've already said something about it in this programming challenge) or Flow in Networks, if someone is interested.
Flow in Networks is really nice technique, it isn't as complex as it sounds and it has wide range of usage. For example, you have one teacher who teaches N students (only 1 at once). You have described when they can attend and you have to build time table. How do you do that without bruteforcing? Correct, you use flow in networks! It has so much use cases and there are another techniques that directly inherit from it.
I could make a post on neural networks, but if you're starting check out cs231n, Stanford. Very good course and explains everything you should need.
Unfortunately not for two weeks at least due to uni exams :(
The title is a bit confusing. I thought this was related to series as in calculus "Sequences and Series" as in summability and convergence or Taylor and MacLaurin series.
I edited it, how's that?
Much better. Thanks!