usgroup

66 ms·

691.
by usgroup 6y ago
It seems about the right time to give it a go. How is the package ecosystem in your experience? Can I expect fairly decent support for numerical optimisation stuff , graphing? Munging xml/Json/etc? Will I find that I have to write
692.
by usgroup 6y ago
If you were trying to motivate someone embedded in Python / scipy / pandas / etc to move to Julia at this time , what would you say?
693.
by usgroup 6y ago
This.
694.
by usgroup 6y ago
See here: https://plato.stanford.edu/entries/epistemology-bayesian/ Section 6.2 gives a very concise list of problems. For further reading have a go at para-consistent logic for more total upheaval of the Jaynesia
695.
by usgroup 6y ago
I didn't allege it wasn't a problem for Frequentists in a similar situation. I'd add that the modelling world is very far from partitioned into Bayesians and Frequentists.
696.
by usgroup 6y ago
If by prior you just mean "I know something about it" or "everything happens in context"; then that fine. But if thats what you mean then a diminishingly small number of events have "priors" which can be expres
697.
by usgroup 6y ago
I think the op is saying, unlike its often presented, by learning Bayes youre not "discovering the truth about how to update beliefs from data". I.e. its not a panacea, its a framework with strengths and weaknesses. Recall, most s
698.
by usgroup 6y ago
I think there is something to be said for when one first discovers Bayes; not least because its often presented like this article as a sort of enlightenment. Perhaps the finest work on the topic is probability theory by Jaynes; if the under
699.
by usgroup 6y ago
I think you're quite right about that which is not to say Bayes can't be very useful but I do agree that it has been elevated to the level of religion.
700.
by usgroup 6y ago
Not sure you need any maths for Kaggle competitions. Learn XGBoost and random forests for regression and classifications and learn how to evaluate the models and some of the tricks for coercing different types of data into them for good pef
701.
by usgroup 6y ago
Applicants should know what the process is; I'm not sure why you'd assume that would be kept from them? It's perfectly normal for hiring / firing sales, traders, and other performance oriented professions. Nothing forces
702.
by usgroup 6y ago
There’s nothing immoral about it if you tell applicants. It’s basically how hiring for sales is done and no one is under any illusion that you won’t keep your job if you don’t perform. The point here is more nuanced than the straw man you’r
703.
by usgroup 6y ago
Sure but you’re bound to say that ... Find me a hiring manager than publicly says anything else. That’s kind of part of the point. I roughly recall google touting ~50% hiring success rate and that’s with a 4 interviewer, 10-15 man interview
704.
by usgroup 6y ago
Well less repeatability issues actually ... since you can think about the counterfactual “what would we lose if they weren’t here” and the actual “what did we gain”. This information is available and you don’t need process to extract it. As
705.
by usgroup 6y ago
It’s a funny cultural thing. Here is a caricature: Hiring managers have a real problem to solve: “how do I hire?”. They don’t have the time to conduct actual experiments and work out what actually relates to better hiring. So they trivialis
706.
by usgroup 6y ago
I think the Olympics is just generally physically bad for the athletes. Runners, weightlifters, wrestlers, boxers, rowers, etc all these folks end up working so close o the edge of sustainable human tolerance and capacity that’s it’s unlike
707.
Blackbaud Ransomware Hack(blackbaud.com)
2 points by usgroup 6y ago|0 comments
708.
by usgroup 6y ago
that's a good point if the two are comparable.
709.
by usgroup 6y ago
It feels anecdotally correct. Consider things like depression as a result of bereavement or child birth (for women). Insomnia , anxiety , etc. So many ways to be “mentally ill” temporarily. That said it does seem that the extreme prevalence
710.
by usgroup 6y ago
I think that they argue that the causality works the other way around. All men are mortal doesn’t mean all mortals are men.
711.
by usgroup 6y ago
Pixel hosted somewhere and then request fingerprinting possibly. Client , IP, country, etc
712.
by usgroup 6y ago
Yes ... you add Bayesian to the thing and it becomes new, principled and superior. It’s been a trend for about 3 decades now.
713.
Ask HN: What advise would you give your T-10 year old self?
1 points by usgroup 6y ago|4 comments
714.
Ask HN: Would you pay for a monthly job market infographic?
2 points by usgroup 6y ago|1 comments
715.
by usgroup 7y ago
I feel lots of people have made the point that over time the emphasis on maths has been reduced such that education would be more useful to more people. I.e structuring the maths syllabus such that it would select for the maths super stars
716.
by usgroup 7y ago
This + when you find someone you like and can understand go consider reading everything else they wrote, look at their references, consider reading them too.
717.
by usgroup 7y ago
I was around when software "engineers" were called "coders". They were then called "developers". To me "engineer" is a bit of an aggrandisement of what is for most of us more akin to plumbing that it
718.
by usgroup 7y ago
Well , it’s deception for personal gain ... To that effect it’s somewhere on that immorality spectrum. Personally, I think one trades morality for gain at the risk of a particularly slippery slope.
719.
by usgroup 7y ago
I feel that the operative difference is between “being” and “seeming” uniformly random.
720.
by usgroup 7y ago
I'd hazard to guess that "information" implies "not uniformly random", and "uniformly random" implies "no information". That the black hole ripples random implies that the input is in part "

More ›