chroem-

25 ms·

211.
by chroem- 5y ago
> Do you have evidence of the contrary? I'm not the one who carries the burden of proof, but I appreciate the sources. Thank you.
212.
by chroem- 5y ago
Do you have data to back that up, or is it conjecture?
213.
by chroem- 5y ago
> The unanswered question here is why Prophet is so popular, and this surely merits a better explanation than I have given. It has been explained by several posters. The goal of Prophet is to make time series accessible to non-experts,
214.
by chroem- 5y ago
The vast majority of time series data that you would encounter in a business setting is seasonal. Most of the datasets he is using are tiny and they haven't had enough history for the seasonality to develop.
215.
by chroem- 5y ago
Yes, it does. I'm not sure what the author is trying to prove, but he's feeding a bunch of non-seasonal data into a highly seasonal modeling technique, and then hurling insults when it doesn't work because that's not wh
216.
by chroem- 5y ago
SoTA in 2021 is the bespoke transformer model you implement yourself based on the idiosyncrasies in your data. Unless a lot of money is on the line, that is overkill for most situations though.
217.
by chroem- 5y ago
The point of Prophet is to make time series accessible to people who aren't experts in time series. Yes, it has some well known failure modes, but using Prophet always beats a static LY "forecast." This article is like comp
218.
by chroem- 5y ago
I have seen Indian posters allude to this before, but I'm not familiar with what the connection is. Could you please elaborate?
219.
by chroem- 5y ago
For business users, unfortunately none on hand at the moment. That being said, one of the biggest issues companies seem to encounter is that forecasting requires logging of at least several years of business operations, especially if the t
220.
by chroem- 5y ago
At my last role I designed the forecasting systems for a major retail chain, which required inventing a new deep learning model architecture for time series problems, due to some peculiar properties of their sales data. I take great pride
221.
by chroem- 5y ago
That's essentially the job role that I'm familiar with, however this new role is everything except for feature and model engineering.
222.
by chroem- 5y ago
The job description was pretty ambiguous, and the interview consisted of pretty standard ML/math questions with some data engineering mixed in. There was one brief offhand comment during the interviews that confused me about the role,
223.
Ask HN: What does an ML Engineer do?
56 points by chroem- 5y ago|41 comments
224.
by chroem- 5y ago
Sure: - consensus - distributed computation - statefulness - peer to peer identity - peer to peer reputation of nodes - resistance against many classic failure modes (e.g. Sybil attacks) - peer to peer source of truth (e.g. Namecoin/DN
225.
by chroem- 5y ago
Kademlia had no consensus mechanism. >I'd argue the biggest breakthrough was the Turing machine Witty, but that's not peer to peer computing.
226.
by chroem- 5y ago
That "glorified linked list" solves a lot of problems in peer to peer computing. In fact, blockchains are arguably the biggest breakthrough that subfield of computer science has ever seen.
227.
by chroem- 5y ago
I don't understand what the positive outcome of this is. If China invades Taiwan they'll be able to hold the rest of the world hostage because of TSMC. It's still a lose-lose situation.
228.
by chroem- 5y ago
A nuclear deterrent only works if you're willing to use it (and basically commit suicide in the process). China is betting that the US isn't willing to risk MAD over Taiwan, and I would say that they're probably right.
229.
by chroem- 5y ago
My point is that it is virtually impossible to establish the concept of wrongdoing without the notion of certain innate rights having been violated. Each person has their belief of what they hold to be human rights. As human beings, it is
230.
by chroem- 5y ago
You just admitted that it was an atrocity however. That implies the people who were killed had their rights violated. Your concept of rights granted by the government to the people is not self-consistent.
231.
by chroem- 5y ago
Flagged for no good reason, so I will post it again, because I think it deserves a reply: You are endorsing human rights abuses by oppressive governments like the Khmer Rouge because that government didn't extend the right to freedom f
232.
by chroem- 5y ago
You honestly believe the concept of human rights is a "rhetorical trick"?
233.
by chroem- 5y ago
Absolutely not, and that's the point. People are only considering this right now because it happens to be politically advantageous in the short term. It is incredibly reckless to subvert rights and institutions for political goals li
234.
by chroem- 5y ago
Hypothetical scenario: Donald Trump's political career miraculously comes back to life in 2024 and he is elected president. He does just as you propose and expands the supreme court with justices of his choosing, and curtails the firs
235.
by chroem- 5y ago
I wonder how swapping the battery packs works. Or do you throw away the entire aircraft after 500 flights? If it's anything like model-scale aircraft, air travel is more taxing on battery packs than automotive, since you have to more
236.
by chroem- 6y ago
For all the pearl clutching about bitcoin, I find it humorous that computer games still consume more energy. [1] [1] https://bitcoinmagazine.com/business/introducing-cbei-a-new-...
237.
by chroem- 6y ago
You're just re-emphasizing that you disagree with the demand preference of people buying GPUs, that they should have used those GPUs for gaming instead of mining cryptocurrency. > The problem arises when manufacturers are unable to
238.
by chroem- 6y ago
Climate change and greed were the motivations you originally provided for attacking bitcoin, but now you claim those things don't matter at all. You're being evasive. The US is a market economy, and like in any other market econo
239.
by chroem- 6y ago
My point is that, if your goals are really to fight climate change and stop greed in finance, then your priorities are ass-backwards at best, and self-defeating at worst. However, this appears to be more about control than actually accompl
240.
by chroem- 6y ago
The focus on gaming is just as arbitrary as the focus on bitcoin, and that's my point. Gaming consumes more energy than bitcoin and yet its economic justification is even less clear. And yes, gamers do cause GPU hardware shortages,

More ›