From one extreme to another: the statistics of extreme events - Jon Keating
Вставка
- Опубліковано 15 бер 2021
- One pleasure of mathematics is its capacity to connect seemingly unconnected problems, & to do it with just a few numbers & symbols. Mountain ranges, family trees, performance at the Olympic Games and Prime Numbers will all be joined by Jon Keating's mathematical thread in this Oxford Mathematics Public Lecture.
Oxford Mathematics Public Lectures are generously supported by XTX Markets. - Наука та технологія
Great talk! It helped me out with a model I'm building. Thanks
from venezuela . What an interesting concept and a great window for research. thank you.
Thanks for explain this topic.
Thank you so much
Great lecture.
useful lecture for those who are interested in applying statistical knowledge to problem-solving in risk management, insurance, financial engineering, banking, and actuarial mathematics and modelling.
The biggest problem faced by the statisticians is that they over-rely on the computational characteristics of the Bell-shaped Normal Distribution Model.
Hence, we saw what happened to such risk models during the GFC of 2008, as most of the incoherent risk measures used by actuaries and quants, underestimated the probabilistic consequences of negative risk-based outcomes aka hazards.
Hello, do you know any papers that discuss this over reliance on the normal distribution model / any other resources? I find it fascinating. Thank you
@@rachelk6628 please do read Nassim Taleb. Thanks
Chaos theory applied.. very good discussion. Considering neuroplasticity and repair/rerouting after stroke.
Good math oxford love you👏👏
If only I had known in high school that maths could be so aesthetic...thank you Jon for this insightful lecture!
55:30 claims that the formula is the same as the family-tree case, but in the family tree case the term that is subtracted had a denominator didn't it?
well explained
I don't really understand the point of the previous comments by I thank you for this lecture
What about a multivariate adjustment where you not only adjust for population, but population density instead, and then add in income inequality, life expectancy, and GDP.
This is mind-blowing!
I wonder how this would apply to econometrics or other fat-tailed phenomena?
s/o from South Africa
Alex,
The answer to your good question is easy: "Completely."
Economics is a lie from top to bottom because the whole thing is built on the notion of the normal distribution -- an arithmetic curiosity which exists in mathematics but not in the real world. An important topic in the discussion of normal distributions is the central tendency of events so distributed. There is no such tendency.
If events in the real world were distributed fractally and without central tendency, all economies would fall off the edges, crash, splatter, or just generally go all to hell every now and then. This is empirically found to happen, so at least we have affirmation of the consequent here. :-)
Adam Smith talks about countervailing forces a lot. Quite right. If one guy tries something, some other guy may very well benefit from doing the opposite. This does not mean they will produce any compromise in the middle: one or the other will sometimes, perhaps often, prevail. Adam Smith is right about how people behave -- which may account for his winsome hope for moral sentiments. Many of his soi-disant followers take quite unjustified liberties in interpreting him.
Long story short: economics is rotten at its core and needs to be completely re-built -- perhaps on the radical assumption that the real world is a factor in what happens in the real world.
This will be slow and difficult because very few people can stand the stresses of operating under such an assumption.
@@TheDavidlloydjones Hello. I find your comment fascinating. I’m wondering if you have any resources (papers/books/etc) related to this topic? Thank you
Beautiful
yes, it is true the gold medals proportional to the population!
no it’s not
Universal Virtue University I am. Founder I was born In Tanzania Trained By Prof Ralph Masenge
Задача а генералах. Теорема о не разреши оси уравнений в радикалах, начиная с 5 порядка.
Statistically speaking...You should watch this video.
Hurst exponent
nice lecture, but Usain Bolt is a Pareto outlier.
Jai shree ram
Jai shree ram
Jai shree ram