Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
iskander
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
33 ms
·
331.
▲
by
iskander
9y ago
That timeline seems to miss that Yann LeCun was already working on ConvNets in 1988. I don't think anyone waited for the Universal Approximation theorem to start building neural architectures, it was just a tangentially interesting mat
332.
▲
by
iskander
9y ago
The anarchist scene has seemed thriving in every city I've lived in.
333.
▲
by
iskander
10y ago
Ah, sorry, misunderstood.
334.
▲
by
iskander
10y ago
>Mars has way more sunlight (much closer to the sun and a much thinner atmosphere) Mercury, Venus, Earth, then Mars. (Mars is further from the sun and gets ~60% of our sunlight)
335.
▲
by
iskander
10y ago
>Body volume is larger, therefore more chances for cancer. Across species that line of thinking actually gives rise to Peto's Paradox: https://en.wikipedia.org/wiki/Peto%27s_paradox
336.
▲
by
iskander
10y ago
It's possible that all the women who answered this survey were trans (since those two categories don't exclude each other).
337.
▲
by
iskander
10y ago
>Merit has little to do with it. Not sure if that's intentional, but the BioNTech series of trials for melanoma been called MERIT (this particular incarnation is Lipo-MERIT).
338.
▲
by
iskander
10y ago
$100k - taxes = ~$65k. Now subtract rent, which in most tech hubs is at least $1500 / month. Now you're left with ~$47k. After the cost of food, transport, and a little travel let's say you still have ~$35k. So each year of w
339.
▲
by
iskander
10y ago
>Where are they now 5 years later? Their loans are mostly paid off and they're in better jobs. This is missing the magical step 2 just before "Profit!". How did your "English major" friends manage to pay off th
340.
▲
by
iskander
10y ago
In theory, Spark lets you seamlessly write parallel computations without sacrificing expressivity. You perform collections-oriented operations (e.g. flatMap, groupBy) and the computation gets magically distributed across a cluster (alongsid
341.
▲
by
iskander
10y ago
> Map-reduce is naturally built into the language I hear this often about both array and functional languages but I think this kind of slogan leads to a misunderstanding of the MapReduce framework (and what's special about it). In a
342.
▲
by
iskander
10y ago
>machine: 16 core 256GB (in all cases: date partition, sym index. all queries in RAM.) ... >all query data is cached in RAM (no disk access). 1) What happens when you run these benchmarks on a machine with only 16GB of RAM? 2) How doe
343.
▲
by
iskander
10y ago
>kdb+ is so fast and powerful you'd think it impossible. I suspect there is a religious/mythological aspect to K/Q's reputation for speed. During the year that I was using Q I often found it to be slower than the equi
344.
▲
by
iskander
11y ago
Hospitals add surprisingly little to health and longevity. The biggest factors are hygiene, exercise, diet, and pollution (including smoking). It's nice to get a speedy by-pass surgery when your heart clogs up, but it's even nicer
345.
▲
by
iskander
11y ago
Would children and retirees also claim basic income?
346.
▲
by
iskander
11y ago
This is actually a lot worse than the results recently reported by Mount Sinai in their comparison of a few dozen patients against two standard lab testing companies. I wonder if Theranos has been trying to clean up their standard tests.
347.
▲
by
iskander
11y ago
I moved from Computer Science into bioinformatics / computational immunology. Bioinformatics is fatally infected by the associated culture of biology. Tools get squirreled away until published in a non-open venue, then eventually dumpe
348.
▲
by
iskander
11y ago
Let me revise that: for solid tumors, chemotherapy can cure people, but typically doesn't. At one extreme, early stage (solid) cancers are primarily cured by surgery. At the other extreme (metastasis), chemotherapy is a terminal roll
349.
▲
by
iskander
11y ago
Q: Did the Dropbox devs working on the second version of Magic Pocket encounter any language stability issues with Rust? Did updates to the core language ever break existing code?
350.
▲
by
iskander
11y ago
Chemotherapy, for most cancers, achieves tumor shrinkage and a temporary extension of life, at the cost of severe life quality degradation. It's a tradeoff that most cancer patients think is worth it. I think that this therapeutic bene
351.
▲
by
iskander
11y ago
Has molecular dynamics had any big success stories? Have new drugs been approved which were initially found by in-silico screening using MD?
352.
▲
by
iskander
11y ago
If you want to write about an exotic other, it's helps to have some actual first-hand knowledge of the people and place to add nuance to your descriptions. Otherwise you're just misrepresenting other people's cultures for ent
353.
▲
by
iskander
11y ago
The top comment by Chris Patil gives an important counterpoint: --- Neuroskeptic’s “shoulders of giants” metaphor betrays an astonishing ignorance of the state of despair around translational psychiatry, and the blogger completely misses th
354.
▲
How much faster is a truncated Singular Value Decomposition?
(blog.explainmydata.com)
1 points
by
iskander
11y ago
|
0 comments
355.
▲
Our Experiences with Flow (Hammer Lab)
(hammerlab.org)
2 points
by
iskander
11y ago
|
0 comments
356.
▲
by
iskander
11y ago
From the summary at the end: >by the spring of the first grade year, teachers rated the TN‐VPK children as less well prepared for school, having poorer work skills, and feeling more negative about school. This was a reversal of the ratin
357.
▲
by
iskander
11y ago
>There's almost certainly a "big data" angle as well, potentially describing the polymorphisms in majorhistocompatability (MHC) molecules between people. Where's the big data angle on MHC polymorphisms? Aren't th
358.
▲
by
iskander
11y ago
Ratings from http://www.ratemyprofessors.com/ShowRatings.jsp?tid=1835732 might illuminate some of the back-story. Lots of unhappy reviews in there: >He went off on so many tangents, nitpicked on random things and when it
359.
▲
by
iskander
11y ago
>many problems are not solved by algebra, but by a method I call “arithmetical reasoning”. Here one merely reasons through the stated operations to find the unknown. No unknown is named, and no equation is set up. The presence of these n
360.
▲
by
iskander
11y ago
It was ~4 years ago on a quad core xeon with ~8GB of memory, using whatever was the most recent version of Q, Python/Numpy and Matlab at the time.
More ›