geebee

50 ms·

391.
by geebee 8y ago
But... isn't that the entire point of splitting your initial training data into a training set and a separate testing set? Why is it better to have an idea of uncertainty from the model itself when you can get the generalization error
392.
by geebee 8y ago
I'm going off a first pass through the paper, but it appears that what this paper shows is that the training error can be 0 on an entirely randomized data set, but the generalization error - the difference between the error on the tes
393.
by geebee 8y ago
Well, right. A binary classification system must assign a 0 or 1. But cross validation or other methods may reveal that it isn't a good fit, just a fit. As mlthoughts pointed out in a different comment, any kind of regression technique
394.
by geebee 8y ago
Thank you for the link. I'll read this paper. I'm hoping to reply but the thread may be stale by the time I do. Right now, my thoughts are: if it is easily fitting, what are the assignment probabilities? Are we getting 90%+, or is
395.
by geebee 8y ago
“Machine learning algorithms will always identify a pattern, even if there is none” Is this partly a problem with interpretation? Let's say I do a binary (supervised) classification with an algorithm that is also capable of assessing p
396.
by geebee 8y ago
The NYTimes discusses this question in the context of ethics, but the hiring process for software engineers, generally speaking, often avoids any discussion of what will be built. I've interviewed at google and a few other similar plac
397.
by geebee 8y ago
I believe that the PE in Industrial engineering works that way (ie., as a professional title designation). I don't think there's anything you can do with a PE in Industrial Engineering that you wouldn't be allowed to do as an
398.
by geebee 8y ago
Is it achievable at scale for all colleges once they all start to make that effort, though? And if so, it it achievable the way Harvey Mudd did it? Harvey Mudd is a very elite college with considerable resources, and it has a relatively sma
399.
by geebee 8y ago
Sadly, misrepresenting is a common political tactic. Remember, "you didn't build that?" President Obama made the argument that businesses do rely on substantial government services that are critical to a healthy commerce sect
400.
by geebee 8y ago
Does he, though? That's a stretch. If someone believes that 20% of the elite computer scientists will be women, why does it follow that he would not have confidence in the women who take his class? For instance, suppose about 20% of hi
401.
by geebee 8y ago
It's worth reading through, I'd recommend it. It doesn't take long.
402.
by geebee 8y ago
I agree. I am disappointed with the headline, though. The headline reads: "A UW professor argued that women don’t want to code. What do women computer scientists have to say?" The sub headline reads: "Six women computer scien
403.
by geebee 8y ago
I read through the responses. It appears all disagree with Stuart Reges. The first few responses disagree, a couple toward the end go farther and discuss the harm that occurs when someone expresses this opinion. One essay does insinuate tha
404.
by geebee 8y ago
I think both questions are relevant. So while I do agree that we should hear from women and men who considered and rejected pursuing computer science, I am also interested in hearing from women who did pursue computer science.
405.
by geebee 8y ago
This raises an interesting question about MOOCs for me. Several people here are objected to violating the terms of a free online course. At what point is it ok to start creating derivative works? I've noticed many of the comments discu
406.
by geebee 8y ago
Interesting. Here's a list of mid career pay by major. Fields that tend to be accredited and value accreditation are certainly present in high places on the list, but variants of mathematics are in comparable places. My guess is that &
407.
by geebee 8y ago
This is interesting, and I think it may explain why so many members of my own family growing up (and now) have gone to University of California schools. This article from the NYTimes really drove it home: Top Colleges are Cheaper Than You T
408.
by geebee 8y ago
There may be some ambiguity in the way you're using the phrase "safe space".
409.
by geebee 8y ago
I agree completely. But for now, that's a pipe dream. If you want to be in this world, resources like this do help. "Cracking the coding interview" is a great study resource, but how do you actually get feedback and practice?
410.
by geebee 8y ago
This is valuable. Thank you for creating and posting it. First, it's an excellent reference for people who don't really understand technical "interviews" and their exam-like nature. Most interviews in other fields do pro
411.
by geebee 8y ago
The RAND institute published a paper supporting this position: It's a bit old now (2003), but I think it still holds. https://www.rand.org/pubs/issue_papers/IP241.html
412.
by geebee 8y ago
Well, for a developer earning over 100K in San Francisco, an additional $20K won't make a difference. Those devs will still be able to eat as much $12 avocado toast as they want, and they still won't have any hope of buying a hous
413.
by geebee 8y ago
Yeah I realized that a little after, I posted an answer to your actual question above
414.
by geebee 8y ago
Oh, my apologies, I understand the question you were asking. You weren't asking for my opinion on how this should work, you were specifically asking about what kind of data is needed. Ok, got it. I'm interested in how viable a car
415.
by geebee 8y ago
That's true. Is the cost of firing really that high, though? And how good are the tech tests at determining that a software developer will be good at working on a product? It seems to me that you could be great at printing all paths in
416.
by geebee 8y ago
I don't have an answer to that question that I'm completely satisfied with, but I can give you my thoughts on it. I'm extremely wary of formal ABA or AMA style licensure. It's not that I oppose it in theory, but requirin
417.
by geebee 8y ago
It would be helpful to see more data on this. I consider the CS degree at a reputable university (doesn't have to be elite, just a proper curriculum) to be very challenging compared to many other majors. And I do know plenty of program
418.
by geebee 8y ago
Other fields also pay well and can draw top talent away from software engineering. Take a look at "US News Best Jobs" and drill down by salary in top paying regions for each job. The information is a roundup of BLS data. https:&#
419.
by geebee 8y ago
Fair point. However, I think the above still holds if you replace “lower prices” with “reduce the growth rate of prices”.
420.
by geebee 8y ago
I think the article raises an important point. I'm skeptical that increasing the supply of housing in San Francisco will lower prices. I do think that San Francisco suppresses a lot of potential business activity through restrictive ho

More ›