throwaway283719

16 ms·

91.
by throwaway283719 12y ago
No. L2 regularization (ridge regression) also has an analytical solution, and L1 regularization (lasso) can be more efficiently solved using the LARS (least-angle regression) algorithm. Gradient descent would work for ridge regression (it
92.
by throwaway283719 12y ago
I get that this is suppose to be an introduction to gradient descent, and that the author chose a simple example to work with to illustrate the method (and explained it very well - kudos!). However, they should probably at least mention tha
93.
by throwaway283719 12y ago
Like I say, I'm not accusing you personally of anything! And I agree with you completely - most people that help, do that because they think that helping people is the right thing to do. I'm saying that there are a whole other (mu
94.
by throwaway283719 12y ago
If you mean "how can you know you can trust the charities on GiveWell", the whole point of GiveWell is to rank charities by how effectively they use their donations to help people. If you mean "how do you know you can trust G
95.
by throwaway283719 12y ago
I have heard a lot of people say that they do this, but I have never seen someone actually do it, nor have I heard a second-hand report of someone doing it. I'm not saying that you're not telling the truth, but I am saying tha
96.
by throwaway283719 12y ago
That's quite a lot of money (assuming you mean USD and not ZAR). Do you have a source for that figure? Is that an average, or is it the largest amount that one person has ever made?
97.
by throwaway283719 12y ago
That's interesting, if it's true. Do you have a source?
98.
by throwaway283719 12y ago
Did you try searching the article for the phrase "Larry Niven"? He is mentioned about halfway down.
99.
by throwaway283719 12y ago
I'm not clear what the lesson is here, unless it's "you should have bought Bitcoin in 2012".
100.
by throwaway283719 12y ago
I started implementing the 'scmutils' library that this book uses in Haskell. I stopped working on it a year or so ago (because life) but it's here if anyone is interested - https://github.com/chris-taylor&#x2
101.
by throwaway283719 12y ago
The point is that humans don't need the context to understand the dialogue.
102.
by throwaway283719 12y ago
I don't know if you're aware, but the site you linked to (The Guardian newspaper) has nothing to do with the BBC.
103.
by throwaway283719 12y ago
If you're going to link to a comic with a joke in it, don't put the punchline in the title of the submission!
104.
by throwaway283719 12y ago
> I would claim it's actually less risky because it's more diversified Yes, I agree (but then again, you could just replicate the index in your stock portfolio...) > owning 100k in stocks would also get you dividends that fu
105.
by throwaway283719 12y ago
Okay, I'll bite. Why is having a naked futures contract in the S&P500 (which costs around $100,000 for the e-mini) any more risky than owning $100,000 of stocks? [0] http://www.cmegroup.com/trading/equity-index
106.
by throwaway283719 12y ago
You say you believe that your ability to get a degree doesn't correlate to anything other than your tolerance to debt or how rich your parents are. But you clearly believe that your 2:2 in Physics (for Americans, a 3.0 GPA) is worth mo
107.
by throwaway283719 12y ago
I'm glad that CloudFlare has a sensible approach to hiring, but sadly I think they will be the exception, rather than the rule.
108.
by throwaway283719 12y ago
He could start by rewriting his resume. This in particular stands out - Accomplishment highlights: * Autodidact. High school dropout. No degree. To almost any employer, that is not an accomplishment, much less a highlight! If you