Bayesian Statistics with Hannah Fry
ฝัง
- เผยแพร่เมื่อ 28 มี.ค. 2019
- You can buy Hannah's great book Hello World from loads of places: www.hannahfry.co.uk/helloworld
Including signed copies on Maths Gear!
mathsgear.co.uk/collections/b...
My book Humble Pi is available signed on Maths Gear, as well as un-signed from all the usuals.
SIGNED mathsgear.co.uk/collections/b...
UNSIGNED www.amazon.co.uk/Humble-Pi-Co...
The history of Thomas Bayes.
www-history.mcs.st-andrews.ac....
All about the theorem named after them.
mathworld.wolfram.com/BayesThe...
Huge thanks to the current residents of the house who tolerated us running around and invading their weekend. All in exchange for two awesome maths books.
Thanks to Ben Sparks for crafting the Geogebra animation of the surface.
CORRECTIONS
- Nothing yet unless you count a slight continuity error with the door in the background. Let me know if you spot anything more serious than that!
Thanks to my Patreon supports who do support these videos and make them possible. Here is a random subset:
Christian Schadewald
Thomas Hodnemyr
Richard Fourie
Maksym Borodin
Laura Dale
Support my channel and I can make more maths videos:
/ standupmaths
Music by Howard Carter
Filming and editing by Trunkman Productions trunkman.co.uk/
Sound production by Peter Doggart
Design by Simon Wright
Runner and assistance by Max
MATT PARKER: Stand-up Mathematician
Website: standupmaths.com/
NEW BOOKk: wwwh.umble-pi.com
Nerdy maths toys: mathsgear.co.uk/ - บันเทิง
"Everytime you're throwing the ball you're getting more information about... " where the table isn't
Ah, the Parker Throw...
Parkesian statistics
I don't know why they didn't just play Battle Ships... its based on Bayesian reasoning!
"where the table isn't"
It's a Parker Square table…
@@ekinteko I believe you and your comment to be genius!!! 😱
I believe this based on my never having thought of this example nor have I seen anyone else explain bayesian statistics with such a sublimely simple example. I, and the others, all use really stupidly complicated examples.
Note, I'm not saying I'm a genius. But I'd bet, at least one of the other mathematicians I've referred to regarding Bayesian statistics, might be.
BUT @Kangal, like Rev. Bayes (who I believe to be a believer) and Pascal (who I'd wager was also a believer), I believe you and your comment to be genius.
If there wasn't a virus going around I'd kiss you. 😘
- "What's the chance you find me here?"
- "It's the first time I've found you here, so I don't know"
I'll use that joke, with your permission...
It’s all yours!
That's a very frequentist approach to chances. After his conversion to Bayesianism, Matt wouldn't have said that ;)
well, we know that it's more than 0%
Chance is about predicting so in fact we actually know that the chance of them meeting is 100% since it happened, therefore has become a fact.
"It's the first time I come here" would have been funnier. :D
Hannah is getting plenty of practice as to being a mum of a toddler here.
This might be the best comment I've ever seen on TH-cam
Ahahahaha 😂😂😂 this REALLY got me, for some reason!
I've watched this video several times and I can't believe I've never noticed this comment. And when I did, I burst out laughing and nearly knocked over my cup of coffee onto a mum with a toddler on the next table (I'm in a cafe)? What's the statistical probability of that (i.e. there being a mum and the toddler while reading a comment about a mum and a [metaphorical? toddler)?
“Its okay, we can do it again 😃”
“How good are you at catch” “I understand the theory” 😂😂😂
My history in experimental physics 😂😂
Matt Parker: Making a pregnant woman repeatedly bend down and pick up his throws, while he is sitting in a comfy chair :P
I didn't notice she was pregnant
Parker chivalry
@@Peter_1986 You mean like being near a staircase?
I didn't even realize she was pregnant before your comment. Now that I looked again it's so obvious...
@@Peter_1986 I don't think they were saying it was dangerous, just a little rude since it's harder to bend down when you're pregnant :)
Schrödinger, Descartes and Bayes walk into a bar. The bartender asks: "Would you gentlemen like a drink?"
Schrödinger says: "Yes and no."
Descartes says: "I don't think so."
Bayes says: "I'm not certain."
didn't get the descartes part
"I doubt therefore I think, I think therefore I am" - René Descartes
Bayes: "Ask us again..."
Descartes: "I think not" *disappears*
I'm going to use this one
"Embrace uncertainty, but quantify it" is my new motto.
This has the same energy as "Take it easy, but take it"
Can I just have Hannah and Matt explain everything to me? The sheer enthusiasm for math on display is infectious.
"I was a certaintist...I don't know."
You don't sound like a very good certaintist.
He was certain that he didn't know!
I know that I was a certainist, but now I'm fairly confident that I am a bayesianist.
I used to be indecisive, but now I'm not so sure.
Bayesian sounds like a colour. I'm not sure what colour.
@@RaglansElectricBaboon Beige-ian.
"And thanks to the owners of the house! Turns out, one of them's a mathematician..."
This is the least surprising statement I think I've ever heard in my life.
"It's all about that Bayes" - Thomas Bayes probably.
There's a paper titled "Drum'n'Bayes". They use Bayesian networks for pattern recognition on music.
"Its all about that Bayes - no table" :-)
I think, you guys may be a little bit biased.
Some Parker throws right there
In fairness to him he did give it a go, even though he wasn't entirely successful
Clearly it was a Parker Square Table. 🙂
6:03 One mathematician tells another mathematician to be more random.
The irony here is delightful.
Micah Long why ironic?
@@user-ky6vw5up9m
I think because you can be either random or not. It's hard to be "more" random. You could increase the sample size of course or the step size, but random is random, it doesn't become more random.
4:10 Hannah's exasperated declaration of "Try" should cause you lot to call and thank your mothers.
I love your profile picture and username! I wish more people were familiar with Berzerk.
Hanna Try
When I was in high school statistics we were talking about Bayesian probability, and one of my classmates asked "Isn't that a black asian?", and my teacher said "no I believe that's blasian" and I will always remember that interaction.
I've never seen Bayes' fundamental insight so amusingly explained. Nor so clearly. Nicely done! (Also: congratulations to Dr Fry on the forthcoming addition to your family!)
I have heard that throwing the ball on the table example before, but actually seeing it being done makes it so much clearer to understand!
Have you read the paper they are referring to? This experiment is discussed in Section II. It's one of those things you have to draw on paper or do in real life to get a proper feel for what Bayes' saying. What I found really interesting is that Rice's motivation for submitting this paper was for it to be used as proof of Intelligent Design. 😮
@@warrenyazzie9975 Actually I'd say that Bayesian view excludes gods alltogether. There's no confidentialy, that you can measure, about exitence of a physical god. There is good examples of a methaphorical gods exitance though. Look for example all churches and temples and what not and ask if the builders had a god in their mind when they did those. But a mtphorical god is like justice: it exists only in deeds and thoughts of men.
Literally bought Hello World an hour before this video was released. Now I have 2 great maths books to get absorbed into. Maths books are like buses, you wait years for a good one and then 2 turn up almost at once!
Do you think Hannah's book is good for actual computer programmers? I would read it if it is more about the ethics and psychology relating to algorithms, but if half of it is trying to explain algorithms in laymen's terms I would not enjoy it.
quote of the century
@@karlkastor It's definitely more about the ethics of the algorithms used in our modern world, and she rarely explains what the algorithms are and how they work (only enough to potentially peak interest so you go off and Google it). It's not a book to read if you want answers and definitions, it's a book that presents questions you didn't even think about asking in the first place. It's terrifying and hilarious, but mostly terrifying; I highly recommend it :).
6:00 "you're actually remarkably consistent"
"Consistency is only a virtue if you're not a screw-up" -Grant Sanderson
Can always count on Grant to come up with the goods!
"As I throw more balls my confidence in my throwing skills lowers asymptotically towards zero."
I wish Hannah Fry had a TH-cam channel
At least we've got "The Curious Cases of Rutherford and Fry"
@@ViatorRex Oh! Is that this Rutherford?!
they should make an audiobook together where they simultaneously read their 2 books...
Oh god the cacophony...
Check out "Numberphile"
Someone should write two books so that they synchronize when read together.
PLEASE I would love it
A video with two hosts where one is sitting down for most of the video while the heavily pregnant woman has to stand the whole time... Fine work, Matt! In all seriousness, this was a lovely video and the playfulness between you two is infectious.
She's the expert, so she's the one that came up with the experiment.
Anybody else find Hannahs voice soothing? Intelligent and beautiful but the voice gets me everytime.
"If you had an infinite number of throws, and you could throw them perfectly randomly..." he would miss the table most of the time.
It's like the old joke (well, I'm old and I was told the joke) of a physicist trying to explain the Monte Carlo method in a bar, so they describe it in terms of throwing darts at a dartboard. Come back next week and the bartender says "welcome back, I loved your explanation. Has your aim gotten any better?"
Parker throw (I had to)
This is the first i've seen or heard of Hannah Fry. I'm already a fan. She is awesome.
It is official YOU DO NEED TO DO MORE WITH BOTH OFF YOU. The interaction (and faces drown) between you guys is hilarius to watch, and we are learning at the same time. Great video.
Greetings from Belgium.
Statistics indicate Hannah Fry will always be a great guest.
these conversation is like two mathematicians flirting each other.
I would have loved to see Matt throw the second ball directly on the first ball and Hanna tell him that the coordinates of the balls are exactly the same.
Great video! :D
They always have so much fun together, that it spreads and reaches us through.
Always love to see Hannah Fry! Loved this one as well.
Love your silly mathloving face! That smirk and side eye gets me every time!
congrats for the baby!
And to the father
Yeah, another person to suffer on this world.
@@LordNezghul oof edgy
@@LordNezghul only Jesus can end our suffering
@@joshuawan7004 lol wtf?
I don't care what the video is about, I read Hannah Fry I click.
She's just lovely!
You should use this uncertainty to save your Parker Square.
Two of my three favourite Mathologers having fun.
Always love videos with Hannah, congrats on her upcoming family too
I know it's been a long time since this came out, but I have to say, the best part of this video is watching two good friends have fun with each other!
Another Parker/Fry video? Ummm, hell yes! Love you two :D
They should write a joint book titled "Tinker, Tailor, Parker, Fry"
@@zelatrix8545 I'm sure they could find two mathematicians with apt names who can be given credit as co-authors.
Excellent video. Thank you Dr Fry and Matt!
Matt and Hannah are a great team, I enjoyed this video very much.
Mr. Parker I absolutely love your videos! And Hannah being in this one makes it even better!
You guys are great, Matt. Love the dynamic between the duo!
Hannah Fry is a treasure
I love this! I'm taking a class on Bayesian statistics now, and I would consider myself a Bayesian vs a Frequentist.
I love this duo. Great chemistry.
The humor is on point. Love the videos
Great science communicators! Science needs more Matts and Hannahs!👩🏫👨🏫
This is the content I subscribed for and I'm definately not disappointed
Apropos of nothing, please thank the owners of that house for letting us glimpse inside. It's beautiful!
Also, congrats to Dr Fry. I genuinely didn't notice until I read these comments.
I heard "try more random" and I couldn't help laughing.
You guys look like you're having so much fun!
I love it~
Am I the only one getting strong Beverly Crusher vibes from Hannah in that jacket?
Well, I am now
Hush Wesley.
I didn’t the first time but just watched this again, and that’s the first thing I thought when I saw Hannah!
Both of you keep math fascinating, and you're both loved for it.
Their duo is killing it! Good match!
I've never seen a video where Hannah isn't pregnant. So make sure you have the correct data before you make conclusions from it.
Try TEDx videos ;)
Really appreciate this because that equation is on my wall as I’m revising statistics for my semester 2 first year exam and this has given me some insight
This was great! Thanks for the maths and the laughs.
8:26 Matt's face when he imagines having an infinite number of throws
Hannah and Matt are a great combo for making math fun!
You 2 are absolutely delightful.
You guys need your own tv show 💜
Wionderful and clear explanation. And, a priceless end skit of sharing about two books.
“Homing in” Hannah.
Swings and Roundabouts.
The two of you are so entertaining. I really love it. 😁
If Hannah, Matt, and Grant (Sanderson) did a video together that would be amazing. Such lovely humans who share their unbridled joy of mathematics make my day.
Bayes rocks! And probably rules too. Thanks for such a fun rundown on the basic approach
Great fun, as always!
This is the best video I have seen this month.
What a beautiful lesson.
Just reading the part of “Hello World” where Bayesian Statistics was discussed, and it reminded me of this video.
This is probably what prompted me to buy the book in the first place!
Took me a while, but I finally got around to reading it.
Gotta pick up Humble Pi next!
This is basically how gradient decent in machine learning works, very fascinating to see how this type of math developed
Bayesian statistics is basically how long distance sniping works, as the spotter tells the sniper how far off he was on the previous shot in order to increase accuracy
Did you see that?!
That table was a parker square!
This is one of my favorite things on youtube
Excellent, excellent, excellent video.
I studied Computer Science at uni, with a little bit of AI.
Bayesian statistics was important, but I never knew where Bayes was coming from. My eyes are well and truly opened.
And of course, many congratulations to Dr. Fry.
Loved the plugs at the end, very nice indeed
„february”
ah, love all those fresh videos
13:29 is art
I just wrote an exam on Bayesian/Belief networks. It's a very interesting topic. They are very useful. Like Hannah's self-driving car example, robots can never know for sure if their sensory inputs are correct, so you need to account for uncertainty by using Bayesian nets. Matt could do a video on Hidden Markov Models, which are the simplest Belief networks to understand.
just brilliant!
It's hilarious how this video manages to explain the concept well even though Matt's terrible throws made it a complete fluke that his guess was close. The first ball *was* in the South of the table, but it could have been slightly North of centre and he'd have still guessed it was that far South because every subsequent throw landed right on the Northern edge.
probably my favourite mathematician and my favourite school of inference in one video.. Matt Parker, you've done it again
Brilliant video!
Finally! All that stuff I did in a previous job makes sense. I could do the computer stuff, I could do the mathematics stuff; but now I think I actually SEE what it's all about.
Great job. I expect no less from this channel, but this one really struck a chord.
😍Hannah
🤩Matt
The wrap-up from 11:00 or so should be a compulsory part of the education system.
@4:09
"I'm trying"
"No! Try not. Do, or do not. There is no try"
OMG two of my favorite people!
Love you guys.
It's magic how the camera on the table can't see the clamp holding it on the table, but can see the corner of the table it's attached to!
@3:15 for those who missed it
It must be some kind of Schrödinger thing where the observation not only collapses the superposition but makes the camera disappear! By the way, is there a Bayesian quantum theory?
Many thanks to the stars of this - Hannah, Matt, the briefly appearing camera operator, and of course Hannah Jr ...
Hannah and Matt in one video? Great start into the weekend!
Congratulations, Hannah!
I love how Dr Fry was deliberately trying to speak softly in the house, and Matt wasn't even going to attempt it 😆.
As always genius and funny 👌
Keep if up Matt and Hannah 👍
I think it is worth watching Matt’s videos more than once sometimes. I thought he was just being silly updating his belief about the house, but I had forgotten that by the end of the video. But watching again (showing it to my niece) I noticed right away what he was doing.
Nice explanations. It will be very useful to me! Tks.
0:45 on the plaque: _"... the basis of most market research and opinion poll techniques"_
while not wrong, that undersells the shit out of it.
Great collab.
and now on Kolmogorov statistics!