The h Bar

vzn

12:57 AM

in theory salon, yesterday, by vzn

FLUID PARADIGM SHIFT 2018½ / Tmachine blog

Physics Meta

4:38 AM

0

Q: How detailed should answers be?

A lot of the best, most interesting questions here on Physics SE (main site) are simple questions: Why does the cream of an oreo stick to one cookie? Why is a steel ball more likely to break glass than a pillow is? They can be answered simply, with basic mechanics, or they can be answered in dep...

Fawad

8:57 AM

“The smaller the planet, the lesser the gravity, and the bigger and taller the creatures will be.”

skull

“The bigger the planet, the greater the gravity, and the smaller and shorter the creatures will be.”

Cows

9:44 AM

Inspired, and looking over some math work I am doing

~dusting it off from a 2 month hiatus.

Emilio Pisanty

9:56 AM

0

Q: Can we have a photon at rest?

Theoretically, Photons don't have rest mass but they behave as material particles in many cases. But is it possible to stop a photon without destroying it ie. Can we slow it down so much to make it almost at rest and if not Why?

quantum-mechanics photons speed-of-light

Surely a dupe?

@JohnRennie @ACuriousMind @Qmechanic

ACuriousMind

@EmilioPisanty physics.stackexchange.com/q/16018/50583, physics.stackexchange.com/q/71639/50583?

John Rennie

@EmilioPisanty HAMMER!! :-)

ACuriousMind

Good to see that the gold dupe privilege is used calmly and deliberately ;)

John Rennie

I always feel uneasy about closing as a duplicate of a question where mine is the accepted answer. It feels a bit like self publicity.

But in this case the OP's comment to the only answer makes it clear that he is asking exactly the same as the dupe I've linked.

Mithrandir24601

@ACuriousMind Don't forget about the many hours of thought and questioning that occurs every time such a privilege is used :P

ACuriousMind

10:05 AM

@johnrennie They are different and just look apparently same — Abhishek 2 mins ago

It seems OP disagrees, @JohnRennie...

John Rennie

@ACuriousMind what do you think? Should I reopen it?

Or ask the OP to clarify what he is asking that isn't covered by the dupe?

ACuriousMind

@JohnRennie No, I think it's on OP to clarily why the two questions are not the same since superficially I'd agree they're duplicates

Emilio Pisanty

10:37 AM

@JohnRennie edit the other dupe into the banner

And its dupe too

I feel they're closer

John Rennie

@EmilioPisanty yes, good idea

bolbteppa

Man, what does $\theta_A \overline{\theta}_B = - \frac{1}{2} \delta_{AB} \overline{\theta}_C \theta_C$ even mean

Immortal Player

Indians here? Is masters in physics a must to take lectureship in Physics (in India)?

John Rennie

10:55 AM

@ImmortalPlayer there are lots of Indian students on the site, but I think they're all undergrads or JEE candidates. I'm not sure we have anyone who knows about staff positions at Indian universities.

In the UK the starting point would be to do a PhD. You may or may not do a masters as preparation for the PhD.

Immortal Player

That would be great. But, in India, it seems that National Eligibility Test needs to be passed. And NET in a subject seems to be allowed only for those who have completed MSc in that particular subject. :(

I don't know whether it is flexible or not.

Thank you for the help @JohnRennie, I will try in other forums.

Slereah

11:35 AM

Morning

skull

hi

Slereah

@bolbteppa are they Grassmann numbers?

bolbteppa

They are, and they represent Majorana spinors, and the index convention is weird

My best guess is that this is talking about the identity
\begin{align}
\overline{\theta}^2 &= \overline{\theta}_{\dot{\alpha}} \overline{\theta}^{\dot{\alpha}} \\
&= \varepsilon^{\dot{\alpha}\dot{\beta}} \overline{\theta}_{\dot{\alpha}} \overline{\theta}_{\dot{\beta}} \\
\varepsilon_{\dot{\alpha}\dot{\gamma}} \overline{\theta}^2 &= \varepsilon_{\dot{\alpha}\dot{\gamma}} \varepsilon^{\dot{\alpha}\dot{\beta}} \overline{\theta}_{\dot{\alpha}} \overline{\theta}_{\dot{\beta}} \\
&= - \varepsilon_{\dot{\gamma} \dot{\alpha}} \varepsilon^{\dot{\alpha}\dot{\beta}} \overline{\theta}_{\dot{\alpha}} \…

2

1:40 PM

Phew

Moved all three of my cathodic screens by hand to the new flat

Secret

IMPLOSION! 40 inch tube TV

►

TV cathode ray tube explodes

►

Exploding Antique TV with Detonating Cord

►

15 Biggest Explosions ever recorded on earth!!

►

skull

2:07 PM

:O

Secret

Top 10 Demolitions Gone Wrong

►

More hilarious

skull

;-D

Semiclassical

2:29 PM

@Secret those are clear examples of comedy = tragedy + distance

Secret

Exactly

vzn

@Semiclassical hey fyi credited you substantially in my latest blog, thx again for all the juicy leads :)

bolbteppa

4

Q: Is there any simple proof of the no-ghost theorem?

Is there any simple proof of the no-ghost theorem in string theory?

string-theory

bounty bounty

Secret

this world seriously don't have enough ghost fields

Semiclassical

@secret youtu.be/USu8vT_tfdw

Secret

2:33 PM

Just imagine how much more craziness can result when unitarity is constantly violated randomly

vzn

lol secret + ghosts ≈ chocolate + peanut butter :P

Secret

Pardon my snarky attitude tonight cause my newsest batch calculation has a near 99% failure rate for the most nontrivial reason

and that frustrates me into eccentricity

11 foot 8 Bridge

The 11 foot 8 Bridge (formally known as the Norfolk Southern–Gregson Street Overpass and nicknamed The Can-Opener) is a railroad bridge in Durham, North Carolina, United States, that has attracted media coverage and popular attention because tall vehicles such as trucks and RVs frequently collide with the unusually low overpass, resulting in damage ranging from RV roof air conditioners being scraped off to entire truck roofs being removed. The 78-year-old bridge along South Gregson Street provides only 11 feet 8 inches (3.56 m) of vertical clearance. It cannot be raised, because nearby railroad...

Semiclassical

@Secret that video reliably makes me snicker

Secret

it does

Semiclassical

2:58 PM

@Secret youtu.be/V3-UugI0JoA

Lozansky

@Semiclassical Lol @1:04

Not the bridge's fault :(

It almost looks like it speeds up after bumping into the curb

Semiclassical

Yeah, I dunno what happened there

That bridge seems to have more issues than just low clearance eg too narrow

Lozansky

They oughta just build a bridge over the bridge

Semiclassical

A bridge over troubled bridge

Lozansky

Haha my thoughts exactly

Semiclassical

3:04 PM

Trouble with that is that some idiot would go too fast and turn that into a stunt ramp

Lozansky

Yeah but think of the compilation videos

Secret

3:29 PM

7239851 T-01_03-00-00a Convergence failure -- run terminated
7239853 T-01_03-00-01a FormBX had a problem
7239855 T-01_03-00-02a Convergence failure -- run terminated
7239857 T-01_03-00-03a Convergence failure -- run terminated
7239859 T-01_03-00-04a Convergence failure -- run terminated
7239861 T-01_03-00-05a Convergence failure -- run terminated
7239863 T-01_03-00-06a Convergence failure -- run terminated
7239865 T-01_03-00-07a Running or sigterm
7239867 T-01_03-00-08a Convergence failure -- run terminated

talk about 50% failure rate

Well at least I knew HOW it fails

Akash. B

physics.stackexchange.com/users/184625/akash-b?tab=reputation

who hates me too much?^^^

skull

don't worry about it, the system will review and reverse "serial down votes"

user187604

@Akash.B how much of reputation have you lost?

skull

if not, talk to a mod

Akash. B

@user187604 check my profile .I am having only 8 reps

seems like someone is specifically targeting me

user187604

3:34 PM

@skull does serial upvotes also get reversed?

Akash. B

reverse?

skull

Yes @user187604

Reverse = undone

Akash. B

oh

user187604

@Akash.B if a specific one person has down voted your questions then it'll reverse I think.

@skull am I right?

skull

^

Yes

Akash. B

3:36 PM

@user187604 if it is through different accounts?

@skull is there any way to get out of the ban

user187604

@Akash.B you can be targeted by one person for any dispute but with no many.

skull

Nope @Akash.B

vzn

← the bridge(s) reminds me of that expr "slow motion train wreck" aka hbar chat room lately o_O :P

user187604

@skull how much the reversal take time?

skull

24 hours

Akash. B

3:38 PM

@skull then what i can do to ask questions?

oh I think i can ask it here

skull

Sure

:-)

user187604

@skull do you recall about SHM?

@skull simple harmonic motion?

skull

$Mathematics$ Mathematics

Associated with Math.SE; for both general discussion & math qu...

7

49

Ask here^

Akash. B

@skull In quantum theory some physicist are taking about parallel world.

user187604

@skull suppose you have a pendulum.

Akash. B

3:43 PM

@skull I couldn't understand those parts

could you please explain it?

user187604

@skull see his enthusiasm. Great for just a class nine student.

skull

Sorry guys, I have to go.

user187604

@skull bye.

skull

Cya

Akash. B

@user187604 oh thank you, at least there is one person in this world who had realized my true potential

skull

3:47 PM

Be nice?

Akash. B

@JohnRennie can you take a look at my question ?

John Rennie

→ 3 messages moved to trash

@Akash.B which question

Akash. B

7 mins ago, by Akash. B

@skull In quantum theory some physicist are taking about parallel world.

6 mins ago, by Akash. B

@skull I couldn't understand those parts

6 mins ago, by Akash. B

could you please explain it?

Secret

And as far we knew, every since the entrance of The Plan a few hours earlier, it get worse, right skull?

user187604

@Akash.B I'll only way to get advanced to read the next level. If you wanna get these questions answered properly then you should have proper theory of pre-level. So while in 9 read class 10. Give olympiads. They are available for your level also.

John Rennie

3:51 PM

@Akash.B there must be a thousand articles on parallel universes out in Googlespace

skull

@Secret yup

Akash. B

@user187604 I had read the 10th class text book

user187604

@Akash.B you can write those theories in a diary and also queries.

@Akash.B give olympiads also to test your knowledge

skull

Gotta go @secret cya

Secret

bye

user187604

3:55 PM

@JohnRennie how do you remember that much theories and derivations and formulas. Do you have a diary?

Akash. B

@user187604 I know my knowledge is to low in physics .So what is the need of Olympiads,refer some books for me to read

John Rennie

@user187604 I think that when you use formulae over and over again they seep into your brain and you remember them without trying.

user187604

@Akash.B abcd gaurang Tandon samjoe they'll help you much. As I've only heard of olympiad.

@JohnRennie agreed I've seen that also.

Akash. B

@user187604 oh i see

user187604

@JohnRennie in UK at what age people get retired?

John Rennie

4:00 PM

@user187604 the retirement age used to be 65, but it has recently gone to 67.

user187604

@JohnRennie you aren't retired then.

John Rennie

I quit full time work a while ago because i had saved enough money that I don't need to work full time. I currently work part time.

user187604

@JohnRennie oh. In UK people are lot dedicated. In india man like you would be busy with family while you are obsessed with physics.

John Rennie

I'm the UK equivalent of a sadhu :-)

Akash. B

@user187604 I think I will be like that

user187604

4:05 PM

@JohnRennie a lot of students in UK believe in PhD right?

@JohnRennie you mean Saint right?

John Rennie

in the UK you only do a phd if you are going to work in research, either at university or in industry

@user187604 probably more like a monk than a saint

user187604

@JohnRennie no body get to be a professor with PhD in UK?

Mithrandir24601

@JohnRennie Bahahahaha!!! :P "What is your advice oh great Dr. Rennie?" "My young padawan, the solution is simple. You just need to eat more"

John Rennie

Anonymous

@JohnRennie Need some halp. Ever since I installed the new keyboard and battery on the laptop it automatically shuts down every 2/3 hours. And I get the message: "The reader is not installed on this machine..."

John Rennie

4:08 PM

Today's lunch :-)

Mithrandir24601

@JohnRennie It has cheese though :/

John Rennie

What's wrong with cheese?

Anonymous

user187604

@JohnRennie have you maintained your picture figure?

Anonymous

This seems to be an issue specific to Latitude: dell.com/community/Laptops-General/…

Mithrandir24601

4:09 PM

@JohnRennie It makes my immune system attack my throat :(

Anonymous

I'm not sure if I should install the Reader manually

Anonymous

Nor do I know the actual source of the error :/

John Rennie

@Blue dell.com/community/Laptops-General/…

@user187604 I weigh about 66 kg and I'm 178cm tall, so that's about spot on.

user187604

@JohnRennie @Blue in h bar we discuss everything except physics!!!

2

John Rennie

@Mithrandir24601 :-(

Mithrandir24601

4:13 PM

@user187604 There was a brief period in the history of the h bar when it was named the h bar(becue) :)

Which is a much better name, in my opinion

John Rennie

@Blue I would probably disable it in the BIOS

Mithrandir24601

::resists temptation to change name of the room::

user187604

@JohnRennie I'm 5 feet 4 inch and I might weight 80kg so there is perhaps conservation of mass and height if the product of them is concerned.

Anonymous

@JohnRennie I could try that, but what changed the BIOS settings in the first place?

John Rennie

@Blue it's possible the bios got reset when you were messing around changing the keyboard, though that seems unlikely. More likely the cable for the smart card reader got dislodged - I think it's the smart card reader that the message is about.

user187604

4:17 PM

@JohnRennie in Paris food is different?

Anonymous

@JohnRennie Ah, that's possible. I did forcibly disengage the keyboard from some of the chips and contacts

John Rennie

@user187604 I was last in Paris 20 years ago when I worked there for a month. I have no idea what the food is like in Paris these days, though I imagine it's pretty cosmopolitan like any large European city.

20 years ago it was very good, but my goodness if you ask for a rare steak there it's still moving when it's served.

You could just ignore it. I doubt it's doing any harm.

user187604

@JohnRennie please add @Blue

Anonymous

According to the video it should have slided out easily, but strangely in my case pulling out the keyboard was a lot of pain. I'm at least happy that it is still alive. :P

John Rennie

If you have a friendly computer repair shop to hand it might be an idea to get them to have a look. Kaumudi got her laptop repaired for a surprisingly low price.

user187604

4:22 PM

@JohnRennie it's great to stay unmarried. No tension no quarrel no demands no child careers. You're leading a great life.

John Rennie

@user187604 there is no perfect life. All the options have advantages and disadvantages. The key is to make the best of whatever life you end up with and concentrate on having a good time.

In my case that mostly involves physics, food and laptops :-)

Speaking of which:

That was dessert :-)

Anonymous

Potatoes?

Anonymous

:P

user187604

@JohnRennie believe me you have the best life. If you'd married then you had to carry your wife's wishes give for child's career. And in last life you have to rely on that child and his wife if you hand your property to them.

Anonymous

@JohnRennie Yeah, I might try that

John Rennie

4:26 PM

@Blue pineapple!

Potato?????

Secret

user187604

@JohnRennie main reason of my wight.

John Rennie

@Secret Lunch and physics combined. What more could a nerd ask for? :-)

user187604

@JohnRennie eat a pineapple and solve problems. That's how you maintain figure?

Secret

indeed

Mithrandir24601

4:28 PM

@JohnRennie Another computer, obviously

Anonymous

@JohnRennie Hehe, the potato fries my mom makes are often shaped like that (though that is a bit darker shade of yellow). I haven't eaten pineapples much

Anonymous

Pineapples taste quite strange

Mithrandir24601

(I'm now up to 2 laptops and 1 desktop :) )

Anonymous

Sometimes a bit too sweet

user187604

@Blue like combination of physics and lunch!

John Rennie

4:29 PM

@Mithrandir24601 I've been messing about upgrading my server with the biggest mofo CPU that would fit in it.

I've been doing quite a bit of video resampling and needed some major CPU power.

I've squeezed in an E3-1280v2 and it is bloody fast! :-)

user187604

@JohnRennie the website you given in your profile isn't developed for four years.

@JohnRennie have you killed the rat?

John Rennie

@user187604 yes, well, I do very little C++ coding or website work these days.

Most of the IT stuff I do is corporate scripting stuff.

Mithrandir24601

@JohnRennie My (OK, technically the Uni's) second laptop is just a small thing (it's designed to be lightweight for portability), but my personal laptop is mildly crazy :)

John Rennie

It's a gaming laptop isn't it? I remember you mentioning it before.

Jake Rose

 Im having a bit of trouble getting the length contraction from the lorentz transformation could somebody give me a hand?

$\delta$

oops

user187604

4:34 PM

@JohnRennie OK bye

John Rennie

Jake, if you put spaces at the beginning of your post it converts it to fixed width font.

Jake Rose

Never knew that!

Fixed with is a beautiful thing

user187604

@JohnRennie good night. From India.

John Rennie

@user187604 good night. See you tomorrow.

Jake Rose

$\Delta x = \gamma (\Delta x'+v\Delta t')$

John Rennie

4:35 PM

Jake, I wrote the definitive answer explaining length contraction. Let me have a quick search for it.

Aha! Found it:

12

A: "Reality" of length contraction in SR

Lorentz contraction is easy to understand once you realise that it is not a contraction at all. Instead it is a rotation and the length of the object, or more precisely its proper length, doesn't change at all. To see this take the usual example of a rod of length $2a$ aligned along the $x$ axis...

Mithrandir24601

@JohnRennie yep :) The screen is amazing, as is the keyboard, which are two things I really care about :) The microphone turned out to be fantastic, once I figured out how best to use it, as well

Jake Rose

Now S' is the moving frame and S is some fixed frame e.g. the lab

(Will read that post after this @JohnRennie)

So we see that $\Delta t'=0$ as both events must happen at the same time in its fram

But that gives us the lenght contraction the wrong way around?

John Rennie

@Mithrandir24601 I've never owned a gaming PC but I have a Dell Precision M6700 that is a real powerhouse. However it weighs a ton and has only a two hour battery life.

I use it when I'm doing serious number crunching on the move.

Mithrandir24601

@JohnRennie Yeah, sounds about right :P

John Rennie

:: grin ::

this just got flagged:

in Maid Café (メイド喫茶), 15 hours ago, by Taisho

Offensive? Should I vote to suspend the OP?

Anonymous

4:40 PM

50% of all anime would be offensive by that standards :P

Anonymous

Or maybe 25%

Mithrandir24601

@JohnRennie That's what I'm currently wondering... The amusing thing is, if it is offensive there, it's offensive in here as well... So...

John Rennie

@JakeRose yes, when you do the calculation it looks as if you get a length increase not a contraction. See my post for why this is.

@Mithrandir24601 I'll be back after my suspension :-)

The flag got cancelled anyway.

Actually if the original post got deleted presumably it would be deleted here too since I only linked it - I didn't copy it.

Mithrandir24601

@JohnRennie Ah, fair does

Loong

@JohnRennie Taisho is a bot.

Anonymous

4:43 PM

@Loong lol

Jake Rose

Great post @JohnRennie You got any good book recommendations for SR over summer?

John Rennie

@JakeRose I've never read a book on SR. It's simple enough that you can just pick it up.

Jake Rose

Yeah thats true

Once youve got the equations down the problems are pretty trivial

Mithrandir24601

@JakeRose Personal experience: SR makes so much more sense when you do GR in third year, so you're better off spending your time doing other stuff (like preparing for second year and relaxing at least a bit)

John Rennie

@JakeRose I think real understanding comes from understanding what the Minkowski metric is and what it means. SR is fundamentally a geometric theory and the metric is the key object you need to understand it.

Jake Rose

4:45 PM

@Mithrandir24601 Cambridge?

@JohnRennie will look out for that in lectures

Anonymous

@Mithrandir24601 Quick question: On page 4 here, any idea what class "$c^{x}$" means for a vector?

Mithrandir24601

@JakeRose Aye, I was at Caius a few years back :)

John Rennie

@JakeRose Just have a flick through these

Anonymous

I guess it is just some categorization of vectors based on some properties

Anonymous

But not sure which

Jake Rose

4:47 PM

@Mithrandir24601 Always love meeting people who know the grind

Anonymous

Couldn't find anything online

Jake Rose

Got my 1st year exams in like 2 weeks but finding it so hard to revise:/

John Rennie

@JakeRose Everything there should be accessible to students with a basic grasp of calculus. No advanced maths needed.

Jake Rose

@JohnRennie Sounds great. Ill have a look after exams!

John Rennie

Anyway, I'm going to repair to my armchair with a glass of beer and this weeks New Scientist. See you all on Monday.

Anonymous

4:49 PM

Goodnight

Mithrandir24601

@JakeRose We did loads of past papers to help us revise. Never seemed to do us any good during the first three years of undergrad, but ultimately ended up working really well

@JohnRennie Enjoy!

Jake Rose

Currently what Im doing

Im just doing questions by topic I dont understand

Mithrandir24601

@Blue I'm struggling to find it :/

John Rennie

@Blue you can download the service manual for you E4310 from Dell, and that might give you a clue as to what has come loose.

Jake Rose

@Mithrandir24601 Im gonna guess, compsci to physics?

Mithrandir24601

4:50 PM

@JakeRose Yep :) We've spoken before, methinks

Anonymous

@Mithrandir24601 Eeeh, I'm sorry

Anonymous

Page 6

Jake Rose

Yeah Im pretty sure we have

Anyway back to the grind!

John Rennie

@Blue the manual is here

Anonymous

4:52 PM

brilliant.org/wiki/feature-vector

Anonymous

"
A set of numeric features can be conveniently described by a feature vector. An example of reaching a two-way classification[clarification needed] from a feature vector (related to the perceptron) consists of calculating the scalar product between the feature vector and a vector of weights, comparing the result with a threshold, and deciding the class based on the comparison.

"

Anonymous

But my question is "How do you know the classes $c^{x}$ beforehand"?

Anonymous

Or maybe we define it for the input vectors

Anonymous

An example of a feature vector you might be familiar with is RGB (red-green-blue) color descriptions. A color can be described by how much red, blue, and green there is in it. A feature vector for this would be color = [R, G, B].

Anonymous

Hmm, this makes more sense

Anonymous

4:55 PM

@Mithrandir24601 It's okay. I think I got it :P

Mithrandir24601

@Blue So, if you take a set of I dunno, things, then split them into categories, $c^x$ is the 'category' of the 'thing' we've labelled as 'x' (give or take actual maths). We know what $c^x$ is because we're at the training stage - we're telling the computer that "we know what categories all these things are, now do your thing and learn". Or at least, that's my guess based on limited knowledge

@Blue Wiki's example is spam/not spam - I'm not sure if this would work as each colour is a multiple of each of R/G/B, so isn't really in any one category. Maybe there are ways round this though?

Anonymous

@Mithrandir24601 Yeah, I'm still confused about how the basis vectors are chosen.

Anonymous

Say my input contains the mapping of some colors with the respective vector

Anonymous

Like $3 \hat{R}+ 25 \hat{G} + 1000 \hat{B}$

Anonymous

But in that case we are defining the basis vectors beforehand

Anonymous

5:00 PM

But how do we decide basis vectors for a random data set

Anonymous

They haven't explained that clearly

Mithrandir24601

@Blue I'm going to hazard a guess that you're thinking about this in too much of a 'run an algorithm to get a definitive solution method' - the example you've got here is a single pixel, while an input would have N pixels and the objective might be to categorise the object contained in the image, which consists of those N pixels, so you know the colours (and locations etc.) of each pixel, but the computer can't easily tell what the picture is actually of

Or, in different words, the computer already knows everything it can about that point, regardless of basis. It's when you get lots of these together that it gets hard for the computer to actually categorise whatever the thing is, but again, I'm really just guessing

Like, a computer can read all the different words in a spam e-mail, but it needs some method of being able to actually tell that the input e-mail is spam

Anonymous

5:15 PM

13

A: What is vector in terms of machine learning

I would think that much of your problem comes because vector is a general term with many uses. In this case, think of it as a list of values or a row in a table. The data structure is a 1-dimensional array; a vector of N elements is an N-dimensional vector, one dimension for each element. For ...

Anonymous

Ah, finally found a good example

Anonymous

"The feature vector is simply one row of input. For instance, in the popular machine learning example of housing price prediction, we might have features (table columns) including a house's year of construction, number of bedrooms, area (m^2), and size of garage (auto capacity). This would give input vectors such as"

Anonymous

[1988, 4, 200, 2]
[2001, 3, 220, 1]

Anonymous

So the basis "year, number,area, size" seems to be pre-defined

Mithrandir24601

@Blue Yeah, you then want to do something that helps put this into a price range classification or somesuch

i.e. Not something trivial like "average area per room"

Anonymous

5:19 PM

@Mithrandir24601 For a mail it would have some vectors like [Moderately Spam, Low Frequency] for some input words or phrases (which is our training data). Based on that it can calculate the "distance" to another new vector a.k.a word or phrase input, and classify that as spam or not spam. Can't directly classify the whole mail as spam or not spam though.

Anonymous

If our training data is large enough then it might be possible to classify a whole mail too (perhaps)

Mithrandir24601

@Blue To some probability, yeah. The point is that you can't do this with a single letter - you need a series of these, which is the feature vector

(I believe)

Anonymous

@Mithrandir24601 I don't understand. A series of what? A feature vector is simply something like $[2,3]$ in the basis $\{\text{Degree of Spam}, \text{Frequency of Spam}\}$ in this example. The collection of feature vectors for every singly-standing word or phrase of your mail would form the feature space of your mail

Emilio Pisanty

why are folks suddenly hating this one?

25

Q: How do grandfather clocks keep going?

How do grandfather clocks keep going? The pendulum is what makes the clock go. However, the pendulum will slow down due to friction. What energy source keeps the pendulum from eventually stopping?

newtonian-mechanics newtonian-gravity time

physics.stackexchange.com/review/close/219653

Mithrandir24601

5:52 PM

@Blue That doesn't make sense - apparently, "In spam-fighting initiatives, features are abundant. They can be IP location, text structure, frequency of certain words, or certain email headers"

So 'degree of spam' would be the classification vector

Anonymous

Degree of spam is a "feature" :P

Anonymous

"In spam-fighting initiatives, features are abundant. They can be IP location, text structure, frequency of certain words, or certain email headers."

Mithrandir24601

as in, it's the thing you want - you don't start knowing things like 'is it likely that this is spam', you start by knowing the e-mail address, the different words used in the e-mail etc.

@Blue What do you mean by 'degree of spam' then?

Anonymous

Your feature vector would be $[1103.9980, \text{blah-blah-blah}, 100, \text{new offer for march}]$ in the basis $\{\text{IP location, text structure, frequency, email-header}\}$. The elements of this basis are called "features".

Mithrandir24601

@Blue Yeah, sure

Anonymous

5:57 PM

@Mithrandir24601 I'm assigning a numerical value for "degree of spam"

Anonymous

That wasn't such a good example though

Anonymous

@Blue This one is better ^

Anonymous

And to each of the feature vectors in my training data set I have predefined a class

Anonymous

Say this $[1103.9980, \text{blah-blah-blah}, 100, \text{new offer for march}]$ could belong to the class "Moderately Spam"

Mithrandir24601

@Blue But what do you mean when you say 'degree of spam'? When I hear someone say 'degree of spam', I take that to mean it's either "likely spam", "possibly spam", "unknown", "possibly not spam", "likely not spam"

Anonymous

6:00 PM

The other possible classes being "Not Spam", "Severely Spam", etc

Anonymous

@Mithrandir24601 Right, right. But you know that would be a much better example of a class $c^{x}$ rather than an example of a feature

Mithrandir24601

@Blue That's exactly my point

Anonymous

@Mithrandir24601 Gotcha. That wasn't a good example, I agree

Anonymous

The newer example is better

Mithrandir24601

@Blue Ahhh, yeah, the new example makes sense

Secret

6:08 PM

0

Q: Is there a fundamental ontology in Physics? What are the fundamental "building blocks" of the universe?

Is there a fundamental ontology in Physics? What are the fundamental "building blocks" of the universe, such that they are not defined in terms of other simpler things?

ontology

lol ontology

we will probably never found the ontology of our universe

Anonymous

@Mithrandir24601 I'm expecting there would be news channel headlines in a few years: "Quantum spam detectors" :P

Anonymous

Stack Overflow would need one for sure

Anonymous

I actually don't have much idea about how much improvement the quantum algorithms provide in terms of complexity (for pattern recognition)

Mithrandir24601

@Blue It might be the usual story of no improvement on the 'complexity class front', but nevertheless get some polynomial improvement or something

dmckee

7:12 PM

@Blue IN the language of machine learning a "feature" is an input parameter.

Anonymous

@dmckee I guess my second example was a better one

bolbteppa

'Regge behavior is a characteristic feature of amplitudes at high energies; resonances govern the properties of amplitudes at low energies... The behavior of the amplitude at high energies, however, cannot be thought of separately from its behavior at low energies...

... The values of the amplitude in these regions are connected with one another by finite energy sum rules (FESR). Consequently, one may describe amplitudes approximately either using Regge poles or using resonances at low energies... a "dual" approach' (dual models, Venenziano stuff)

Cows

8:15 PM

quassi-trivial corrections from last night calculations

tahkevin Live Stream

►

Anonymous

8:44 PM

@Mithrandir24601 Silly question: How is $(\langle b|\otimes \langle a|).(|a\rangle \otimes |b\rangle)$ defined? (i.e. inner product of two vectors in a tensor product vector space)

Anonymous

Can it be expressed in terms of $\langle a|b\rangle$ ?

Anonymous

Oh and also for $(\langle b|\otimes \langle a|).(|b\rangle \otimes |a\rangle)$ what would be it?

Mithrandir24601

@Blue it looks like $\left|\left\langle a\vert b\right\rangle\right|^2$

Anonymous

Anonymous

I'm confused now

Mithrandir24601

8:51 PM

@Blue And this would just be $\left< b\vert b\right>\left< a\vert a\right>$

Anonymous

Okay, but reason?

Mithrandir24601

@Blue ::grumble:: people do this sometimes when writing down things like density matrices

Anonymous

Is it simply convention or am I missing something?

Anonymous

I mean as long as conjugate symmetry, linearity in first argument and positive definiteness is satisfied both should be fine

Anonymous

I guess

Mithrandir24601

8:59 PM

@Blue So, let's say that you've got 2 Paulis, $I$ and $X$: $\left(I\otimes X\right)^T = I^T\otimes X^T = I\otimes X\neq X^T\otimes I^T = X\otimes I$

Mathematics

Transcript for

$Mathematics$ Mathematics