Scea91

23 ms·

211.
by Scea91 8y ago
It is even worse in Europe where you are often required to have Master's degree before you can even start a PhD. People in my country usually start PhD. at 25 and take at least 6 years to finish, because the universities use them as ch
212.
by Scea91 8y ago
Of course, there is a difference of how much ahead. If you would have invested few months later you could have come much more ahead.
213.
by Scea91 8y ago
I am not sure how much they see the border with Syria as a concern or an opportunity to expand. Isn't Turkey currently occupying Syrian land?
214.
by Scea91 8y ago
The context should be provided in the Vox article but it is not. Is there any article where all of the offensive tweets are available in their context?
215.
by Scea91 8y ago
I am still waiting for flying bicycles. Also, why do you mention only vehicles? 87 % of ALL the energy we use comes from fossil fuels (coal + gas + oil).
216.
by Scea91 8y ago
Another problem is energy density as I said. How much solar energy can you harvest from one square kilometer? I am so surprised that people just handwave this basic fundamental limits away and don't realize that we would have to build
217.
by Scea91 8y ago
Exactly, we haven't achieved that much actually. Most of the technology we use for harvesting energy is not much further from what people could imagine in 19th century. Turbine was invented in the 19th century. I really suggest watchin
218.
by Scea91 8y ago
I do not expect that we will be able to move past fossil fuels in the next 100 years. There is simply no other source of energy with as much density and as convenient as oil. The only think we know that can rival oil is nuclear energy but i
219.
by Scea91 8y ago
What are you talking about? WAIS-R arithmetic score is an observable entity.
220.
by Scea91 8y ago
Technology might have had a negative impact on their families but I believe that the alternative would be even worse for them. Without technology they would probably be conquered in war and speaking German now.
221.
by Scea91 8y ago
26 - Walking 5 meters on slackline.
222.
by Scea91 8y ago
https://www.hotelolsanka.cz/en/sports/swimming-pool/
223.
by Scea91 8y ago
Swimming pool I occasionally visit in Prague uses less chlorine and has installed UV lamps in the pipes. The water quality is great. I hope this gets more widespread.
224.
by Scea91 8y ago
> So, yes, comparing things to gigantic trees can make things tough. But I thought that was the beauty of boosting and building up smaller trees. Most of them are usually more interpretable than you might expect. Regarding your last para
225.
by Scea91 8y ago
https://distill.pub/2018/building-blocks/ This seems pretty good to me. And the nets are not exactly as small as I meant. I work with random forests and build forest which have more than 80000 nodes per tree. Othe
226.
by Scea91 8y ago
I think interpretability is much more correlated with model size than model type. Small neural net is much more interpretable than decision trees with thousands of nodes.
227.
by Scea91 8y ago
For IE at that time there were obvious superior alternatives. There are no such alternatives to Chrome today, so I do not expect any downfall.
228.
by Scea91 8y ago
I wonder if this is an American thing, just as calling even mediocre stuff awesome. In my small European country there are no such pressures. I once left a 2 star review because we had to wait 15 minutes in a restaurant before we could orde
229.
by Scea91 8y ago
I always wanted to try Airbnb, but everytime I searched for accomodation it didn't seem to offer better deals than I could find on booking.com or elsewhere.
230.
by Scea91 8y ago
Not sure, but doesn't Messenger support E2E as opt-in?
231.
by Scea91 8y ago
You would have to define more formally what you mean by one piece of information and even then I think there are counterexamples to your argument. For example, If I wanted to hire a doctor I would prefer to know if the applicant had medical
232.
by Scea91 8y ago
> its new right-at-the-end placement is still prominent for people who read the whole document (as employees should) or who skip to the end. I really wonder what is the percentage of employees that read their company's code of condu
233.
by Scea91 8y ago
True exponentials exist very rarely. What we are used to call exponential is often at most quadratic and even then the trend usually exists only for some limited time window. I wouldn't dare to make prediction about mankind 1000 years
234.
by Scea91 8y ago
It would be great if people feared climate change the same way they fear AI takeover.
235.
by Scea91 8y ago
Only partly. Spark's RDD API which is still used quite heavily requires external serializer. Most of the time you would use Kryo, but it does not fork for objects larger than 2 GB or sometimes for custom classes that are not explicitly
236.
by Scea91 8y ago
His newsletter is actually the only newsletter that I get weekly and skim it regularly. I agree with other commenters that sometimes the material is weak and more like "Do these 5 things that successful people (bankers, investors, entr
237.
by Scea91 8y ago
At least half of my colleagues have degrees in mathematics.
238.
by Scea91 8y ago
You would be surprised how the devil always emerges in the details when you move to real applications from academic exercises and datasets. I definitely use a lot of statistics daily. I do not need to compute integrals by hand or construct
239.
by Scea91 8y ago
> Ah, I was under the impression that RFs choose from a subset of features, not just one feature. You are correct. For classification the usual rule of thumb is to select square root of the number of features.
240.
by Scea91 8y ago
They get the answer by being linked to the question they duplicated. The only problem is if the question is not exact duplicate, but that doesn't happen very often with such easy questions.

More ›