xapata

9 ms·

61.
by xapata 3y ago
True, the networking causes trouble. I usually rely on a communication layer that addresses those troubles. A good message queue makes the two paradigms quite similar. Or something like Dask ( https://www.dask.org/ ). Havi
62.
by xapata 3y ago
How is coordinating between different machines any different than coordinating between different processes? A multiprocessing implementation is a good prototype for a distributed implementation.
63.
by xapata 3y ago
One weird trick ... There's some fox and hedgehog analogy I've never understood.
64.
by xapata 3y ago
Depends. Does Numba help you?
65.
by xapata 3y ago
The audience is rescued from the sad moments by letters to Ndugu. I remember a couple times being on the edge of tears, only to cheer up when I heard, "Dear Ndugu."
66.
by xapata 3y ago
No, I don't think I can make any inferences about psychiatry from those articles. My Bayesian update is the opposite of yours: because the evidence presented was weak, I'm more skeptical of any relationship between psychiatry and
67.
by xapata 3y ago
Thanks. My first reaction is that this one is much better. Larger sample, better methods. Again, I hate the arbitrary bucketing. There's no benefit to creating age buckets, as if there's some magical change that happens at 40,
68.
by xapata 3y ago
Did you accidentally paste the same link twice? Looking at the "Medical Specialty and the Incidence of Divorce" article, my first thought on reading "Scores on these [psychological characteristic] scales were grouped accordin
69.
by xapata 3y ago
The milk doesn't need to be that hot. Warm is fine, and if the bottle is that hot, I'd worry that the milk is too hot to drink.
70.
by xapata 3y ago
So far, glass bottles have been fine for my kid. I was guessing the commenter was referring to the expense of buying glass bottles, because they cost about double, and that "literal nightmare" was hyperbole.
71.
by xapata 3y ago
Looks like my hypothesis is supported by the evidence: On page 4, you can see their regression results. In fact, they don't estimate the statistical significance of any occupation as a factor, not even the law enforcement occupation t
72.
by xapata 3y ago
> This is a preview of subscription content, access via your institution. Unfortunately, since I can't view the article, I can only guess that they evaluated a large number of professions. Since the abstract focused on law enforcem
73.
by xapata 3y ago
I was thinking of something more communal/distributed. WikiLeaks wanted centralization.
74.
by xapata 3y ago
The way to fight back is ubiquitous public monitoring of the state. We should understand that the state is always watching, but the state should also be aware that all its actions will be public knowledge. Note the increased awareness brou
75.
by xapata 3y ago
What's the goal of preventing use of these data? Are there alternative methods that could achieve the same goal?
76.
by xapata 3y ago
> for a lot of Ml usecases Have you forgotten the No Free Lunch Theorem?
77.
by xapata 3y ago
First person: I, me, us Second person: You That's usually called perspective, not tone.
78.
by xapata 3y ago
Or being taught informally by a colleague or a friend. I am curious which is most prevalent!
79.
by xapata 3y ago
Interesting. "Great person" is tough to design.
80.
by xapata 3y ago
Clusters make a nice story, but you're going about it backwards. What are you trying to predict? The scientific method suggests we create a hypothesis first, then design an experiment (or identify a natural experiment), then collect d
81.
by xapata 3y ago
> Would the world be any different if they didn't try to hide the data? Also no. I think they're asserting that the world would have been different. I find it plausible, if not certain.
82.
by xapata 3y ago
Not all people who do great work are great people. In fact, I suspect a negative correlation.
83.
by xapata 3y ago
Sure, but that's why universities like to ask for unrestricted donations.
84.
by xapata 3y ago
Salaries and other compensation can be considered part of the overhead of managing the money and the institution.
85.
by xapata 3y ago
It's a hedge fund that operates a university for the tax advantages.
86.
by xapata 3y ago
Yes, but it's reasonable to note such things instead of simply derogating the remark.
87.
by xapata 3y ago
It's clever, but like (almost) all such things, doesn't always apply.
88.
by xapata 3y ago
An appeal to authority? I suppose it'd be better if you named the article's author or linked to it.
89.
by xapata 3y ago
jedberg doesn't seem like a naïve person, glancing at his work history. Perhaps you could explain why you think this opinion is naïve?
90.
by xapata 3y ago
You have to remember that with spam, they're looking for gullible people, and the marginal cost of the spam activity is extremely low. To some extent, the Nigerian prince scam works not in spite of, but because it's nonsense.

More ›