« first day (953 days earlier)      last day (4366 days later) » 

hi
is there anyone ?
?
10:23
yes
how are you
busy :/
It's weird. Most of the time when I read Asaf's comments the image of an angry mob of chickens pops up in my mind. Same goes for Asaf's comments on meta, if not even more.
On an unrelated note: it's been chucking down snow here. Can't remember seeing so much snow in mid-March.
10:40
@Ilya Hi!
@MichaelGreinecker Hi, how are you?
@MattN. same here, we have a Christmas weather
@Ilya Good, good. Bit cold though, it started snowing again.
I wonder, if it's a present from Siberia
so sad - we had a really warm last week
@MichaelGreinecker I shall finally watch this movie
@MichaelGreinecker: anyway, I had a question to you, related to MDP
10:45
@Ilya It is very odd, but great for people who like classic rock n roll
@Ilya What about MDP?
@MichaelGreinecker suppose we have an MDP over Borel state and action spaces in discrete time.
AFAIK, there are two ways to make this model complete: either assume continuity conditions (semi-continuous model)
or allow universally measurable stuff
i think that is the usual approach
I'm in favor of the 2nd one due to the generality and nice introduction in Bertsekas and Shreve
now, whenever we have a Borel measurable stationary policy, MDP becomes a Markov Chain
usual one, on a Borel space
can we say that the same holds if we fix a universally measurable stationary policy?
it won't be a classical MC, I guess...
10:48
Hm. The policy is not the problem but the solution. Do you want to know whether stationary problems have stationary solutions?
no-no
the point is that, I can assure some optimality of stationary policies, but those are not for sure Borel-measurable
user19161
Hey @ilya!
suppose, there is an inequality I that holds over any MC
thus, it holds under any fixed stationary (Borel?) policy over an MDP
thus, it holds over the optimal policy as well
Some almost-everywhere-inequality?
10:52
everywhere. More precisely,
$$
|\mathsf E^\pi[U(h)] - \mathsf E^\pi[V(h)]|\leq\delta \tag{I}
$$
where $h$ is the complete history (=element of the sample space), and $U$ and $V$ are different utilities
say, I can solve the problem for $V$ and I would like to use it to approximate the solution for $U$
the point is that $(I)$ holds for Markov Chain, and thus for any Borel stationary policy $\pi$
even in case Borel stationary are $\epsilon$-optimal, I have $(I)$ for the optimal policies
but say I don't have the $\epsilon$-optimality
that's why I wonder, which kind of process is it, when I fix a stationary non-Borel policy
any ideas? @Michael
So the question reduces to something like: Is the orbit of a MC induced by a universally measurable function almost surely equal to some Borel function induced MC? I guess not,otherwise we should get Borel optimal policies.
@MichaelGreinecker hm, perhaps - I will bear it in my mind
btw: this is a really cool introduction to relation between optimization universal measurability: jstor.org/…
11:00
@MichaelGreinecker thanks, I'm still not very comfortable with analytic sets and the stuff
@Michael: btw! you may help me with my literature search, perhaps
@Ilya I can feel you. That stuff is wild.
@Ilya I have to leave, I'm hosting a guest here today. I'll be around tomorrow. bye
@MichaelGreinecker ok :) good luck - I'll try to catch you tomorrow in chat
11:36
@Ethan, that's Goldbachs theorem not Euler
the answer is totally wrong......
user19161
@caveman Hey, how is your revision?
and has been accepted
im not doing revision yet
how do you even accept such a wrong answer
user19161
@caveman Just comment on the post.
11:52
I posted an answer
even though he's answering a different quesiton, he doesn't even do that right..
user19161
@caveman You can downvote the answer if it is wrong to you.
it's not wrong "to me"
this isn't a subjective thing about feelings or whatever
12:25
I always judge the correctness of the answers to my stackexchange questions based on my feelings.
12:43
The questioner has that right.
12:56
Top-10 Reasons you might be an MSE addict!

10. You check your MSE stats a LOT. You occasionally get up in the middle of the night and sneak a peak.

9. Your significant other suspects you are having an affair with MSE. Even when you’re alone with your special person, you do find yourself thinking what other postings might be needing your insightful inputs right then…

8. You “mental solve” while driving or on the train, and sometimes even when you are alone in the shower.

7. You filter everything through your MSE approach and start giving it up- and down-votes, even watching a movie,with
9
...you $\Huge\text{think}$ they are your friends even though you've never even met them!
user19161
@κρανίοπεριπολία Why not?
user19161
It's not as if people you do meet in person will stick with you through thick and thin.
user19161
But I am lucky enough to have some very close friends who will do that.
user19161
@Amzoti Hey Mr Regards!
13:07
@jasper you are gray is everything fine ?
user19161
@DominicMichaelis My colour is just what I like to see at that moment. You cannot interpret anything from it!
Haven't you heard, people can pretend to be your friend and it's harder to detect it on-line.
user19161
@DominicMichaelis Anyway, thanks for asking. I will email you again.
user19161
@κρανίοπεριπολία I just wanna point out that people pretend to be your friend in real life too. There's not much diff really.
... it's harder to detect it on-line
user19161
13:10
There are also different levels of friendship, and different levels of trust.
user19161
And yes, I have met enough assholes online and offline to know there are all kinds of people in this world.
what do you mean with "again" you never emailed me before ^^
user19161
@DominicMichaelis Yeah, that's right. I was wondering if that was correct, linguistically.
user19161
I am not sure now.
user19161
@DominicMichaelis I see you got two downvotes recently!
yeah but no idea for what
@DominicMichaelis you can see them in your reputation change along with the post.
I see where i got the downvotes
but i don't see a reason for it
ohh, that is an old problem.
13:14
@DominicMichaelis people don't always give reasons
unfortunately
it can be some people trying to get their reputation to a multiple of five.
user19161
Also, over time, friends can become enemies, and enemies can become friends.
@JasperLoy I'll probably just stick with JSchlather I've seen another Jacob here and there.
user19161
@MattN. Yes, I think the same as you. Asaf is quite comically angry it seems.
oh i still can get such a lot reputation today
i hope there come some nice questions for me
user19161
13:19
Although I am happy that I got many votes for the trig question, I am unhappy that I got less votes than the other answer.
user19161
That is because in my arrogant opinion, my solution is the simplest.
user19161
Anyway, I gave everyone upvotes on that post.
user19161
It is also very weird that I got 3 votes today for a question I answered a few days ago, even though there is no bumping.
user19161
Sometimes, this happens because there is a duplicate or link.
13:26
bumping ?
user19161
Bumping means maybe a new answer or edit was made.
user19161
This pushes the post to the main page again.
user19161
It is then said to be bumped.
user19161
I don't upvote another person more than 5 times a day.
user19161
Otherwise, it will be removed by the script or the mods, or be suspect for sockpupetting.
13:29
I love this guy's name. It represents all of us
user19161
@Ethereal I don't use WA.
user19161
The only math software I use is the calculator, and also TeX.
@JasperLoy I like your dishonesty. :-P
user19161
@Ethereal I really don't use WA. I don't tell lies.
@JasperLoy It's fine... we lie sometimes :-)
user19161
13:31
@Ethereal In this case, I really don't use WA.
@JasperLoy Okay.
i don't use wolfram alpha i always use mathematica :D
Hi all
Hi!
Just wondering... does anyone here know about generating functions? :)
13:42
not much
ok
I was just trying to solve a recurrence, and was running into trouble.
@anorton what recursion?
oops.
2
Q: Generating Functions: Solving a Second-Order Recurrence

anortonI'm self-studying generating functions (using GeneratingFunctionology as a text). I came across this programming problem, which I immediately recognized as a modification of the Fibonacci sequence. I wanted to place my newly found generating function techniques to work, so I tried to solve the ...

It's really a fibonacci sequence, but with larger rabbit litters... :)
$$a_{n+2} = a_{n+1} + ka_{n}$$
where $k$ is a positive integer.
but where do generating functions come in?
To solve it...
13:45
(I can see that one could solve it using them, but it does seem like an obvious way to go)
I know there exist formulas for solving without generating functions, but I wanted to do it with them, just for practice
since $k$ is constant, this is essentially like a second order linear ODE
ahh, ok
been a while since I did anything with this
I know the book "analytic combinatorics" has a nice treatment of what happens to the generating functions when they satisfy certain recurrence relations
Hmm....
but I don't recall the author, and I am not even sure if it has been published
I think I'll look into that book.
Would it be Sedgewick?
Sedgewick and Flajolet published a book by that title
13:49
looks like the one
Ok. I have a copy of that... I'll look there.
I read a pre-publication copy quite a few years ago as part of a course on this
well, read part of it I mean
Cool. That's pretty awesome.
14:02
Don't you just hate it when you make a simple (high school) algebra mistake??? :\ I made a mistake factoring in that recurrence...
Happy pi day :D
user19161
@Washu Why is it pi day?
user19161
OIC.
@JasperLoy going to make my pie when i get back home :)
4 hours ago, by κρανίοπεριπολία
$\Huge\text{Happy $\pi$ Day}$
14:18
@κρανίοπεριπολία :D
:D
24
A: How many digits of $\pi$ are currently known?

κρανίοπεριπολίαThe current record is ten trillion and fifty digits. The last known digit is 1 .

@κρανίοπεριπολία wow that's a lot
that won't fit in my pie
make a bigger pie
it is only $\pi$-day if you write the date in that illogical american way
the real $\pi$-day is of course april 31st
14:41
@κρανίοπεριπολία I am not sure that question is worth asking anymore, since now we don't have to calculate the digits serially and anyone can calculate any specific digit from $\pi$ using the formulaes with no requirement to calculate the previous digits.
@TobiasKildetoft I love you for saying that.
next step is to convince people to add that day to april of course
3
Another pi day is written in ISO format as far as digits allow, 3141/5/9
@κρανίοπεριπολία That is a cool phone jacket/sleeve.
there is no 31 st april
14:49
7 mins ago, by Tobias Kildetoft
next step is to convince people to add that day to april of course
oh
i guess february don't mind to lose anthoer day
@JayeshBadwaik Yep, notice how you do math as soon as you wake-up?
@κρανίοπεριπολία Haha. Yes. :-)
Math is everywhere
mh still got some pancake better
14:51
kicking my butt for not learning properly...
will make the rest
but one day i will :)
@DominicMichaelis more like taking out a day from august to prevent double 31 in a row.
Will make counting easier for sure for many.
i can't count anyway
Basically have 13 months of 28 days every year, and add five day extra week every fourth year. :-)
14:59
that would be better
Good day!
What does mean a wedge product of differential form and current? $\omega \wedge [X]$ for example
@κρανίοπεριπολία hi
"meromorphic (1,0)-form $\omega$ is called weakly holomorphic if $\omega \wedge [X]$ is a closed $\bar{\partial}$-current"
aaaaa, got it
user19161
Ladies and gentlemen, I have capped for the day, thank you for your support.
15:07
congratulations
i still need more than 100 ...
I am still at 0 for today
@JasperLoy WHAT? How did you do that so fast?
I've never capped... :\
user19161
@anorton That is slow. Capping at GTM 0100 is fast.
Have you done that before, too? :D
user19161
Maybe a bit more than that, yes.
15:09
the fastet cap i made was at 8:00 GTM
user19161
Around GTM 0400.
user19161
Tomorrow I need to do something important, wish me luck.
user19161
Also, 640 more points to attain 20k...
15:17
yeah good luck
user19161
I was thinking, if today is pi day, which day is e day?
@JasperLoy the 71st of February, what else? :D
user19161
@anorton I was thinking, maybe 27 Jan.
Ok... in that instance, I guess the American system of writing dates is a bit off...
@JasperLoy: Where are you?
user19161
15:25
@GarbageCollector I am in the world.
user19161
My actual location shall not be revealed.
@JasperLoy Oh, you changed your blue screen. I usually locate you by searching the blue box.
user19161
@GarbageCollector Yes, I like to see different colours.
user19161
@GarbageCollector The most reliable way is the user number.
@JasperLoy That is correct. Btw, I got 3 down votes for tex.stackexchange.com/questions/6676/…
15:29
without a reason as far as i can see
user19161
@GarbageCollector And what are the downvotes for?
Skullpatrol
@JasperLoy Maybe from the TikZ fanatics who don't like me to criticize the PGF manual.
user19161
@GarbageCollector I think each has its good points, let people use what they like.
@κρανίοπεριπολία You are a raiders fan, right?
Then lament with me the treachery that has happened.
For Wes Welker is now in your division, which means he is no longer on my team
user19161
What is Raiders? Is it baseball club?
15:33
Football (american football)
user19161
Ah, football, not soccer.
user19161
Of course, I call soccer football.
The Broncos just got about four times more scary.
@JasperLoy and @:8511208 you look nearly similar.
user19161
@GarbageCollector What is 6275?
15:34
i don't know if it is still snowing or snowing again
@JasperLoy The skeleton.
Seriously, and they signed Rodgers-Cromartie? What the fuck?
@Arkamis We ain't afraid of no DONKEYS HHHHHEEE HAAAWWW
You should be
Two years ago that team was starting Kyle Orton and people were putting up billboards begging for Tebow
Now they've got Peyton Manning throwing to Wes Welker, and just signed one of the best CBs in the league. This is like bizzaro world. If you told 2011 me that this was what the future looks like, I would have had you committed to an insane asylum.
There is a rumor that Chucky may return to the Raiders....
... or at least the NFL.
15:41
That's an annual rumor. I'm shocked it hasn't happened yet
I posted a question on the Sports.SE about the Raiders wanna see?
yeah
Is the entire text of the question just "WHYYYYYYYYYY??????????"
I found it!
@Arkamis Good job Sherlock.
15:46
heh
@GarbageCollector i like the way the pgf manual is written f
The answer is basically "Al Davis."
He got old, and insisted on having his old crusty hands in the soup
He is the leader of the Dark Side
@Arkamis Big Al
@DominicMichaelis But unnecessary information makes it verbose.
@GarbageCollector I did not find it too verbose. :-/
15:50
@JayeshBadwaik The talk about Karl's students is verbose (for me). :D
let me google verbose
i did find them funny
@DominicMichaelis Karl's students should not be explained too much. :D
@Arkamis Do you have any idea about PSU and purdue computer science programs for undergraduate?
@JayeshBadwaik Not really.
I don't know much about CS in general.
okay, no problems. :-)
15:55
Both schools have generally good reputations overall, though
My grandfather did his electrical engineering degree at Purdue, way back in the day.
Had a couple of friends who recently graduated there, but they were Civil Engineering and Nuclear Engineering, respectively.
You're a Pats fan right?
Yeah
The tuck-rule will burn in every Raider fans heart for the rest of our collective lives....
Legit call.
16:02
Man, you might have gotten a bad break there
But that's nothing.
David Tyree -- and the five missed holding calls on that same play. The ball sailing through Rodney Harrison's hands. Welker's drop in 2012. Bill Buckner. Aaron Boone.
Really, being a Boston fan is a life of suffering.
2
user19161
@arkamis Why did you change your name?
@JasperLoy Because I wanted to distance a post from my real name somewhat.
user19161
@Arkamis I see.
I made a fairly scathing analysis of the current way statistics are used in science.
user19161
16:06
@Arkamis So what?
Didn't want that to come up on a Google search. Many of the people who review my work have backgrounds in stats.
user19161
Oh, so work related.
user19161
Anyway, I think I won't go to the other room anymore. Too many people have hurt me there.
Which room?
user19161
Sometimes, I don't know if it is intentional. I can't guess people's thoughts.
user19161
16:13
Something could be said jokingly, or something could be said otherwise.
@JasperLoy I hurt you?
@Arkamis What is the deal with declaring major after two years? Is it possible that one does not get to select a major because it is very competitive? (in reference to Penn State college of Engineering)
user19161
@GarbageCollector Not you.
user19161
Hey @peter how are your classes?
16:24
why is there no very low quality option for closing a question
user19161
@DominicMichaelis Aren't the other options enough?
it would fit for the homework questions where no effort is done
user19161
@Arkamis The Eng room.
@DominicMichaelis Because voting to close for "very low quality" would rule you out as a "bad person" to put it nicely.
@JasperLoy They are a bit strange there.
16:27
@PeterTamaroff this has been posted here before.
user19161
@κρανίοπεριπολία I never flag people in chat though, and I try to look at things in multiple ways before deciding what people mean.
@JasperLoy But that is JUST you ...
user19161
@κρανίοπεριπολία But it is true that everthing said has multiple interpretations.
@PeterTamaroff Hahaha.
16:29
@JasperLoy Yes, context is everything.
user19161
But there were two occasions in the past where I was really upset in that room. They were quite long ago.
Being upset in that room makes no difference to them I have found.
user19161
So I think I won't go there anymore, QED. Maybe they wanna get rid of me too.
user19161
After all, I am just a pineapple to them.
16:30
Perhaps.
You are a banana here.
bananas are no bad fruits
I enjoy them.
@JasperLoy It's better to be a big fish in a small pond, than a small fish in a big pond.
user19161
@κρανίοπεριπολία It's better to be a banana than a pineapple, QED.
user19161
@κρανίοπεριπολία Yes, that is true.
16:39
@JasperLoy You are the $\huge\text{BIG BANANA}$ around here pal.
user19161
@κρανίοπεριπολία Anyway, I hope Charlie is fine.
Me too :(
@JasperLoy Like they say "if you can't stand the heat stay out of the kitchen."
user19161
@κρανίοπεριπολία Do you feel hurt by some people in there too?
@JasperLoy Yes, but I always try to protect my feelings on the internet.
i have no feelings, i am a man :D
do you know sledge hammer ?
16:49
Peter Gabriel?
oh its in german wait a secodn
I $\Huge\text{think}$ I got him?
Not much of a thinker...
@JayeshBadwaik Some universities make people declare after two years because the first two years are generally all the same.
16:59
@Arkamis As a Raider Fan I am obligated to show you Patriot Fans this

« first day (953 days earlier)      last day (4366 days later) »