nonbel

21 ms·

151.
by nonbel 8y ago
Thanks, I think I added the link to the paper while you were writing this. From skimming it I can't tell the answer to my question though. How do they know what appears "black" in the image is really black vs. relatively blac
152.
by nonbel 8y ago
It doesn't sound like they just snapped a picture. The one guy says they used "supercomputers" for 6 months to get the image. Sunspots look black relative to the rest of the sun but are actually very bright. Could this be the
153.
by nonbel 8y ago
No. I am talking about decreasing the revenue for a government funded business then later paying for it indirectly via taxes. The price of the service (train ticket) does not change, but you are still paying. That is what makes it obfuscate
154.
by nonbel 8y ago
> "Yes, the cost is obfuscated, just like every marketing cost of every commercial company is obfuscated." The cost of marketing for a normal business is built into the price of the product. How is that obfuscated?
155.
by nonbel 8y ago
So you are not sure if they are receiving tax dollars or not, but think they probably do. It doesn't really matter if it is for a specific purpose or not, if they have less revenue from other sources they will need larger subsidies. W
156.
by nonbel 8y ago
Do dutch state railway companies not receive funds from taxes?
157.
by nonbel 8y ago
More and more people are down-voting reality on this site. It will be interesting to see what this correlates with.
158.
by nonbel 8y ago
It wasn't either of those things. It was only possible due to the development of novel ink: https://www.bl.uk/treasures/gutenberg/ink.html
159.
by nonbel 8y ago
Thinking more on this... Those are either 90% or 95% confidence intervals (couldn't tell from the recent paper). So that would mean if you looked at 100 galaxies you would expect 10 (or 5) to fall outside the range if the model was co
160.
by nonbel 8y ago
The point is this "window" isn't narrow at all once you allow dark matter. Practically any theory plus 95% "dark matter" can match the data.
161.
by nonbel 8y ago
>"[dark matter] seems to make up about 95% of the galactic mass...crushingly narrow passages between observation and tests of GR" GR only matches observation because you added in 20x more stuff that is undetectable other than a
162.
by nonbel 8y ago
I didn't look at the other one but NGC1052-DF2 is consistent with MOND: Original sigma = 7.8 (+5.2/-2.2) New sigma = 8.5 (+2.3/-3.1) "Crude" MOND sigma = 14 (+4/-4) https:/&#x2F
163.
by nonbel 8y ago
MMT = next step to hyper-inflation of the USD Do people really disagree with this?
164.
by nonbel 8y ago
Neither of those block spoofed numbers, both work by creating a blacklist.
165.
by nonbel 8y ago
I'll take your word for that. But the problem is that bad regulation has created a downside for providers to block spam calls so they continue not to. I would easily switch providers tomorrow if one of the features was "no spoof
166.
by nonbel 8y ago
Nope, turns out it is some sort of crappy regulation preventing providers from stopping the robocallers from spoofing numbers in the US: >"The FCC last year authorized voice service providers to block more types of calls in which th
167.
by nonbel 8y ago
>"Science would benefit tremendously if this was the approach." This is just the original idea behind science as developed in the renaissance. The resistance to checking each others work is a recent phenomenon. Edit to add a so
168.
by nonbel 8y ago
>"assuming there's double-blind competitions" So even the author does not know if they are human or AI?
169.
by nonbel 8y ago
> "~80,000 people died from the disease" This number sounds highly questionable to say the least: https://www.cdc.gov/flu/about/burden/how-cdc-estimates.htm With no way to compare it to a gold st
170.
by nonbel 8y ago
Not saying I disagree (I tend to favor rational behavior myself), but this sounds like a dogma. What evidence is there that only acting rationally is "better"?
171.
by nonbel 8y ago
That is true. I didn't read carefully enough.
172.
by nonbel 8y ago
> "we as laymen should refrain from coming to conclusions until after a thorough investigation has been completed" People say this, but why? It seems to go against human nature. People love to speculate based on limited inform
173.
by nonbel 8y ago
>"A strong example is a classifier that distinguishes between wolves and dogs. When you look at which pixels were the most impactful in 'Wolf' predictions, you can see that it is actually pixels of snow that are leading to
174.
by nonbel 8y ago
Yes, precisely. It is a common practice in medical research as well.
175.
by nonbel 8y ago
The thing is the ML models don't try to be interpretable. Instead people try to find models that make correct predictions. AFAICT, that is also the appropriate role for simpler methods like linear regression, unless you have reason to
176.
by nonbel 8y ago
Check this: https://news.ycombinator.com/item?id=19353204
177.
by nonbel 8y ago
> "the models are 100% explainable" In my experience this is largely illusory. People think they understand what the model is saying but forget that everything is based on assuming the model is a correct description of reality.
178.
by nonbel 8y ago
>"but they were doing "model explainability" before it was even a thing!" Do you have an example?
179.
by nonbel 8y ago
>"What features in the data did the model think are most important?" This is going to be more misleading than informative, since the feature importance hierarchy can change drastically based on what other features are included.
180.
by nonbel 8y ago
Thanks. Any insight into how they justify grouping in unknown poisonings with those due to common medications?

More ›