Allen Downey - A future of data science

Рет қаралды 17,906

Posit PBC

Күн бұрын

Пікірлер: 24

@frictionless Ай бұрын

Great and interesting talk! Thanks for sharing :)

@MKultragram Ай бұрын

Great speaker. Thanks for such outstanding resources Allen!

@kolawoleadeshina2781 Ай бұрын

This is one of the best things i came across today. Thank you

@respanol1970 Ай бұрын

Amazing talk. Thank you!!

@mudbone7706 Ай бұрын

Interesting talk. Regarding global warming outlook, recent studies have significantly increased the probability of an AMOC collapse (a climate tipping point) within the next century. Not sure if speaker was aware of this.

@zebmason6530 Ай бұрын

I really enjoyed this. Will have to have a look at the duration of pregnancy stats as it looks like the same distribution as adult heights. Despite the best efforts of a famous Belgian that people are trying to write out of the Index of history (who managed to get the normal distribution of heights as a standard textbook example) the outliers show that the distribution is not normal as pointed out by Mandlebrot in The Misbehavior of Markets.

@carvalhoribeiro Ай бұрын

Great. Thanks for sharing this

@dennisfidha Ай бұрын

The Elements of Data Science book is not free. Otherwise great presentation!

@TorbjørnSmedsund Ай бұрын

"What's there to be happy about? Job's not finished. Job finished? Nah, I don't think so." (Kobe, 2009)

@pookiepats Ай бұрын

Kobe was just so cool, the champion of hard work & dedication - RIP 🪦🐍

@zebmason6530 Ай бұрын

ENIAC wasn't the first programmable electronic computer, Colossus was.

@MrAllenDowney Ай бұрын

I think I remembered to include all of the qualifiers: ENIAC was the first programmable, electronic, general-purpose digital computer. Colossus was not general-purpose.

@samimidost6587 Ай бұрын

I was shocked by a professor's reduction of people's happiness to just two factors: life expectancy and the number of deaths in natural disasters... Yes, negative news certainly impacts people's feelings of unhappiness, but there are many other factors at play. For example: the increasing difficulty for young people to find good jobs, the deterioration in income distribution, and the rise of individualism.

@roon1sicunt Ай бұрын

A few more: the global extinction rate; land clearing; fresh water degredation and depletion; housing affordability; suicide rates. All these are rising and continue to do so. Deaths by natural disasters should of course be lower now than ever. Its inconceivable to think otherwise IMO. A bit of a silly one. Very data scientist of him.

@brad349miller Ай бұрын

My world view results: UN facts I scored 90% correctly. Latin America I scored 100% correctly. I am 35 and only hold a highschool diploma. I failed reading and writing from kindergarten to second grade and when I was tested again in 5th grade I had college level reading comprehension. In highschool I failed English all four years[.] I am a statistical nightmare.

@wuzixiao1 Ай бұрын

Like the point of thinking like statistics. But not agree data scientists are different from combining statistics and computer science

@letstalkaboutmath2121 Ай бұрын

I didn't like a lot the judgments of the statistical field. I don't think that statistics miss the train. I don't even understand well the claim honestly. There's a whole bunch of statistics branches that count on the support of the computer power. And honestly who ask for which test to use doesn't know enough of statistics. A good statistician doesn't have a match between setting and test, it has the knowledge of a general framework that will guide it naturally to every test needed, even some not yet formalized with a name like a t, chi or fisher test

@thomasspeidel3308 Ай бұрын

Says, statistics missed the boat. Then proceeds to demonstrate the bootstrap, one of the main breakthroughs of modern statistics on a tool that was developed by statisticians. I enjoyed the talk, but the speaker seems to misunderstand the field of statistics. Perhaps, he could have argued about the need to reform stat education.

@letstalkaboutmath2121 Ай бұрын

@thomasspeidel3308 or even bayesian statistics that only in the last 70 years could thrive thanks tò computer Power, although was formalized even before than classical frequentist statistics

@haraldurkarlsson1147 17 күн бұрын

I still don't understand the definition "Data Scientist". Data is not Science. Science is what you do with the data. Therefore you need a domain - e.g. in the sciences.