rm999

31 ms·

241.
by rm999 10y ago
Completely agreed on negotiating without the "current salary" perspective. I've always requested to keep my current salary secret, but I've never met a hiring manager/HR recruiter that's ok with this. I suppose
242.
by rm999 10y ago
Would Catalyst in Spark SQL count? http://people.csail.mit.edu/matei/papers/2015/sigmod_spark_s...
243.
by rm999 10y ago
People do talk about it a lot; it's often referred to in an optimistic way: "getting your vr legs", instead of a negative manner. That said, the early adopters of VR are self-selecting and probably the least likely to get nau
244.
by rm999 10y ago
I agree machine learning isn't going anywhere - and yes the big companies are going to solve increasingly complex problems using increasingly complex methods using their data and resources. The trend I'm seeing is medium sized com
245.
by rm999 10y ago
This article mirrors some huge frustrations I've had in recent years as a long-time lover and pusher of machine learning. I spoke to a Watson booth employee for a few minutes at a machine learning conference a couple years ago, and alm
246.
by rm999 10y ago
>high density housing, pollution Low density housing and sprawl are actually worse for the environment btw. You have a clear bias against urban living. That's your opinion and that's totally ok, but the cultural trend is defini
247.
by rm999 10y ago
Oh yeah definitely, I meant just for the resume filtering stage.
248.
by rm999 10y ago
If an applicant has less than a couple years experience, it's fair to expect/ask for GPA. If they have a lot of independent projects and a clear passion, I'll often excuse a low/missing GPA. GPA is about as flawed as alm
249.
by rm999 10y ago
I've been saying this for a couple of years, but if I were into real estate investing I'd pour money into rental properties in secondary and tertiary cities. I lived in Baltimore 10 years ago and was visiting last weekend. Neighbo
250.
by rm999 10y ago
I think it requires strong, perfectionist, top-down leadership. See for example Steve Jobs or James Cameron. They aren't the most popular bosses, but they create amazing things by working their people hard and watching everything very
251.
by rm999 10y ago
Looks at spectral clustering. Spectral methods usually take the distances between points to generate a graph, then operate directly on that graph. If you already have a graph/network, you can use those methods directly on it.
252.
by rm999 10y ago
Technology isn't just an industry - it's a way of doing things and is becoming ubiquitous. When people think "tech" they may think of sharing status updates or searching for a restaurant. What about driverless trucks (tr
253.
by rm999 10y ago
The current state of hiring reflects two facts: 1. A single bad hire can do immense damage to a team and organization. I never fully appreciated this until I saw the consequences of bad hires, and as a hiring manager it's one of my big
254.
by rm999 10y ago
>Acceleration is needed for training -- not running the models themselves This isn't true. Running neural networks (including CNNs) can be computationally and power intensive, and lends itself to the vector operations of GPUs, FPGAs
255.
by rm999 10y ago
Ignoring the clickbait article, HIIT (high intensity interval training) is incredible. I did this routine 3-4 times a week for a couple months: http://www.fitocracy.com/knowledge/dr-layne-nortons-best-dam... and got my
256.
by rm999 10y ago
In general the tech community has gotten a lot softer (online and IRL) over time as the field has become more professional and mainstream. In my experience the stereotypical tech jerk with open contempt for everyone - see for example Nick B
257.
by rm999 10y ago
Your comment is mostly addressing what I refer to as "data skews, misuse of stats, and external factors", and is what the article is about. Yes, those are huge issues, but they have little to do with data size. I'm confused w
258.
by rm999 10y ago
I think "big" data is being confused with "ubiquitous" data in this article. Larger datasets will always lead to more statistical confidence in a conclusion you make from the data. The article does a great job of explain
259.
by rm999 10y ago
>I would not recommend ... it is an attempt to influence votes. Are you accusing him of breaking a rule or a spirit of a rule? Because your comment reads like that. To me, this is an unequivocally useful service, and those imaginary inte
260.
by rm999 10y ago
>Someone has to write those libraries Yes, but the discussion is about the "extremely common" case of prototyping, which definitely should not require library building, and usually should not require veering much from establish
261.
by rm999 10y ago
>It's extremely common to prototype in a language like python or R to be able to bang out a draft idea before diving into making it fast. I think what's far more common is prototyping something in Python or R, and then realizin
262.
by rm999 11y ago
There's some confusion here, I do believe my explanation means [25, 8, 5, 3, 3] has an h-index of 3.
263.
by rm999 11y ago
Isn't that what I described? That user has 2 submissions with at least 2 points each. I also think that's a fairly sql-esque way to calculate h-index. I can't think of how else I would do it.
264.
by rm999 11y ago
For anyone wondering, I had to read the wikipedia page and sql code to fully understand what the h-index means in this table: an h-index of n means the user has n submissions (stories, jobs, or polls) with a score of at least n. So, ColinWr
265.
by rm999 11y ago
He did start the book to sort of mock the company; from his NPR interview: >when I first sold the book and start to write it, it was meant to be sort of a modern-day "Office Space ... I wanted to write just a funny story about being
266.
by rm999 11y ago
Could it be this? https://news.ycombinator.com/item?id=9432949
267.
by rm999 11y ago
I did the analysis about 9 months ago for my team when the 980 ti's came out, and the AWS pricing was expensive (we built a server that paid for itself in 2 weeks compared to a g2.8xlarge). This is largely because the 980 ti is actuall
268.
by rm999 11y ago
My biased opinion: VR is going to be huge, but it won't look that way for at least a few years while things mature (especially when 90% of the content is just improving games that could have existed on a 2D monitor). Just off the top o
269.
by rm999 11y ago
The brain needs to form structure before any sort of reinforcement learning is even possible. This structure comes from an "unsupervised" process we don't understand well.
270.
by rm999 11y ago
Probably from Best Buy, for those wondering: https://news.ycombinator.com/item?id=434850 I like to put things into perspective: their revenue back then was around 40B a year, so this represented about a 0.3/40= 0.75% i

More ›