Introduction to Markov Chains, Part 2

[UoCvodka]

Good as Sal (at Khan Academy)

[Ramshobraja]

Man, this made it so much easier than my retarded textbook.

I get this! *Bro hug*

[patrickJMT]

@Ramshobraja ha, glad i could help

[abcdjames]

Thank you!

[GreyBillie]

Thank you very very much!

[alexglew]

Thank you very much, I now understand markov chains! /"that matrix feel"

[formchoi2190]

why can't my professors be as clear as you

[gangesh8]

thank you very much sir

[jessemaurais]

It's pretty neat how the Markov chain converges like that. I was just learning about convergent sequences in R^n the other day for an economics class, but this is the most interesting example I've seen so far.

[ODESrap]

greta thanx man, it is increadible:) i have not understood my prof.

[ODESrap]

Great thanx man, it is increadible, i did not understand my prof...

[siraekabut]

Love your videos!! I will watch all of them asap!!

[MahsaRS9]

Thank you, this was very thourough and easy to understand :)

[MrHav1k]

You sir are the greatest math teacher of all time!

[pebre79]

Great examples and very well explained! Thanks for posting this. Can you post Part 3?

[Belem809]

thank you so much!

[Faraz70]

where is the tech definition? good job

[EntropyDivideByZero]

Found Part 1. YouTube for iPhone did not find it earlier!

[EntropyDivideByZero]

Where is part 1?

[PelicanGodOfJupiter]

You. Freaking. LIFESAVER.

[iaplminute]

I understand you much more than kahn academy

[topconpro9161]

lool really good video... but i really wanted to know your definition of markov chains.. which got cut off at the end.. again..

[kbilal89]

knowledge is the only wealth, the more you spend, the larger it becomes

[ampiyush]

dude.. whr's the next part of this series......

[GlompoFie]

How do you solve a markov chain when it tells you the exact number of people using brand A and not using brand A. Like if it says that a survey was done that said 235 people use brand A and 190 people do not use A. And then you are asked to find the probability matrix and transition matrix?

[Kuki1810]

@GlompoFie u find % of tht.. im guessing here.... 190 out of 235 is 80.85% ppl using brand A and etc...

[SlickPictures]

Thank you so much

[Mattprole]

I've waited for a year without a conclusion to my Markov chains. The comment section is a lonely place.

[8uti26]

Great Job. I read my university script and went over some websites and I really need examples and calculations to understand math not only formulas and results. Your video has helped me to understand in minutes where I sometimes have to sit hours to solve the puzzle.

I wonder if most of the lecturers and teachers have a desire to tease and torture their students or if they are just unable to clearly explain a topic.

You do a superb job!!!

[Bogdan1980]

Thanks for your video. I wanted to ask if the 1 week term is arbitrarily set right away or is that a standard way to determine the initial period? Would anything change if the initial period were 1 month?

Thanks

Bogdan

[sarahkrogers]

Really super, cheers!

[freedomchinaify]

that's so cool, thank you for your tut video. now I have a very good understsnding of markov chain!

[ramiropatricio]

Really outstanding way to explain Markov Model, transform something complex into something simple

Thanks for your video

Do you have some explanation for HMM?

Thanks again

Ramiro

[sibiasa]

i love u patrick-from spongebob

[smicha7]

You are a very talented teacher. You should joint your forces with KhanAcademy. There should be a world school about every subject. Keep working. Great job!

[patrickJMT]

@smicha7 i do not think he has any desire to ' join forces ', but you can tell him that.

[joeprajae]

@smicha7 totally agree

[crosthom]

thankyou so much for your help!



[Attillae]

Great work. But where is part 3 please

[horrabletypoe]

Is it possible to determine a time at which the system reaches equilibrium? I assume you'd need to find a convergent limit, but since turning the process into a function isn't particularly intuitive, I have no idea how to go about it.

[beethz]

i never understand these things in lectures, but you have saved me.

[checkxfile]

Great. In my job I have encountered Markov Chain Monte Carlo. But I never really understand the details. Where is the part 3?

[weaverjwc]

Where is part 3

[ScourgeOfGod3]

Great job!!! I learned a lot and it was very informative. You saved my ass 'cause I was going to be screwed tomorrow if I didn't understand this. Thanks again!

[rsxdarkxsin]

wait.... what about after week 2? if you multiply .9 by .9 by .2, and add .8 of .3 of .3, you end up with 23.4%. So if your supposed to multiply the first branch, and add the multiplication of the second branch, each week less and less people use brand A, which doesn't make sense. explain?

[andyroo2540]

man thank you so much. Really gave me the head start on Markov chains that my teacher couldnt be bothered to give

[JSA19882007]

wheres part 3?

[patrickJMT]

@JSA19882007 it is still in the ether... or my head.

it was just going to be a 'mathy' summary of the ideas involved... i need to make it soon

[nooret]

@patrickJMT nice vid. But i'm also hoping to watch part 3 very soon :D

[gbriceno]

@patrickJMT Nice job on the overview and sample problem.....getting part3 up would be a nice wrapper for the explanation...

[jepz168]

Great help so far! I also can't find part 3

[aswad877]

thanks dude. i'm trying to apply this to DNA which is a bit confusing..but you defo. helped

[Halfaraisin]

What happened to part 3 ?

[paskavatta]

I think these are the very best videos about Markov chains basics. What I like to see is Hidden Markov Models (HMM). HMM is more used in voice recognition where state sequence is not well known. Patrick, do you have plans about HMMs?

[joshhough00]

is part 3 posted?? i cant seem to find it :(

[Xelastic]

Thank you very much!

Nice review :))

used to do this at high school :))

[mrlovespread2006]

you would do markov chains in high school? that's impossible... you wouldn't have enough time to complete the prerequisites of a year of calculus and a probability course to do stochastic processes.

[Xelastic]

@mrlovespread2006

well we did something analogous (transition matrices .. etc) to what is shown in this video

but it was not calculus based or whatever ...

[paulio2293]

this is just the bare basics, these are time homogeneous markov chains (discrete state space), they get incredibly complex when doing problems that involve history dependence (semi markov models), and also inhomogeneous jump processes with duration dependence, kolmogorov equations etc. the higher level study of markov chains is atleast 3rd or 4th year probability theory. Basic markov chains dont involve any calculus.

[canaan1967]

Fascinating

[kkringel]

Thanks!! What if some buy orange juice daily and others weekly? There would be a frequency distribution for that from the market research. In this lesson the transition probabilty distribution is probability that they will switch/stay within a certain period of time (no matter how many times they buy within that period). If probability dist was based on the event of buying, and distribution of buying frequencies was known, how would you get A's % market share vs. time?

[juliomoralesm]

Estos videos como ayudan, son la Neta

[sdfdsv]

I say it again: great topic, so thank you patrick!

[gero1990]

Have u upload the third part??

[amjPeace]

Thank you for the lesson, Patrick. It feels good to be able to understand what it all means instead of just going through the motions!

[khalid97899789]

thanx patrick 4 the hard work

[mwitney1917]

@khalid97899789 yeah how good is he? I came online to watch porn and I ended up watching this instead lol