Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
bravura
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
71 ms
·
931.
▲
by
bravura
14y ago
Recommendation: Reach out to my colleague James Bergstra, and build out automatic hyperparameter selection. This will make your offering work off-the-shelf, which is what is necessary for it to see wider adoption. Why? The real pain in the
932.
▲
by
bravura
14y ago
It's worth noting that buying power != standard of living. For example, it is quite possible to have a high standard of living with little money in Manhattan. This is because: * the subway runs 24 hours * there are numerous inexpensive rest
933.
▲
by
bravura
14y ago
I've only been following Svbtle with passive interest, so I cannot recall or comment on how his responses may have been off-putting. I would like to explain my hypothesis as to why he jumped on the fact that people copied the theme . Consi
934.
▲
Anonymous Leaks Video of Steubenville HSers Joking About Raping a Teenager
(jezebel.com)
1 points
by
bravura
14y ago
|
0 comments
935.
▲
by
bravura
14y ago
The other comments about theories of how this works are insightful, and I encourage you to read them. I also wanted to point out the Facebook acquisition of mysterious Kuala Lampur startup Octazen: http://www.quora.com/Why-did-Facebook-acq
936.
▲
by
bravura
14y ago
I have no idea what this has to do with lean startup. Lean startup is not about: technical efficiency, a distributed workforce, or keeping your burn low. "The Lean Startup relies on validated learning, scientific experimentation, and iterat
937.
▲
by
bravura
14y ago
The number of layers is one of the factors controlling model complexity. Interesting, the number of units in the hidden layers isn't as important as the size of the weights of the hidden units. This is a famous result from the 1990's. H
938.
▲
by
bravura
14y ago
You are incorrect. The ability to control for structural risk in random forests and boosting is comparable to that of SVMs. (For the audience, "structural risk" = "model complexity". Structural risk can cause overfitting. Hence, Occam's raz
939.
▲
by
bravura
14y ago
I think Workflowy has an amazing approach to this problem (hierarchical, collapsable tree) and I simply wish they added date functionality. Right now, I can't use it for TODO, and use it only for brainstorming. For TODO, I now use getflow.c
940.
▲
by
bravura
14y ago
He's taking a systematic approach to de-risking areas of the business that he doesn't feel comfortable with. This is the same thing you do when you jump into a large code-base. He's resisting the temptation to use sales skill on himself .
941.
▲
by
bravura
14y ago
I love this recipe for a one-dish-meal in a rice-cooker: White Chicken Chili http://www.food.com/recipe/crock-pot-white-chicken-chili-114... It's like someone decided that chili was too colorful, and stripped it of all tones. It's also de
942.
▲
by
bravura
14y ago
Three more tips for important talks that you want to nail : 1) Go through your slides and remove as much math/code as possible, until you have a minimal amount of math/code. Now, go through your slides and remove as much text as possible
943.
▲
by
bravura
14y ago
No. Big data has a formal definition, but one that most people ignore: When the size of the data becomes part of the problem. (An O'Reilly author said this once, I forget his name.) For example, physicists in the 80's who had 10s of MBs of
944.
▲
by
bravura
14y ago
Actually this isn't a bad model: If you use software to easily create syndicates, you could raise a few months, assuming you keep your burn at two guys in a garage. Kickstarter meets angel.co If you put money in the company, it's completely
945.
▲
by
bravura
14y ago
pmarca: I think you're onto something here. A large micro-investment portfolio. Anything in this portfolio is sort of a tip of the hat from a16z, a micro-endorsement that messages: "we've got our eyes on this company."
946.
▲
by
bravura
14y ago
What is the correct way to use angel list if you're not currently fundraising? How should one engage on the site? What is the strategy in deciding who to follow? How do you attract people to follow you?
947.
▲
by
bravura
14y ago
Does that mean that recipients can legitimately tell the press and other investors that a16z is one of their investors? Assuming that the recipient can keep a straight face? EDIT: Perhaps having a16z as the only VC in your seed is a very ba
948.
▲
by
bravura
14y ago
Does anyone have insight into how painful it is for non-technical people to query their data warehouses? I'm building a tool that allows business people and non-technical analysts to query their data warehouses using natural language . (
949.
▲
by
bravura
14y ago
Could someone please explain the risks of plugging into a random USB device? I see a handful of people alluding to risks (viz. "USB Condom") but I don't understand precisely what the issue is. On Windows, presumably there's autorun.inf whic
950.
▲
by
bravura
14y ago
Is it possible for non-investors (i.e. small business owners) to avoid taxes using strategies like a "Cayman Islands mail drop"? If so, I propose this HACK to solve the tax issue: Create a site that disseminates information on how small bus
951.
▲
by
bravura
14y ago
I would like to use this, but can I place it on a subdirectory of my domain? It doesn't appear that I can. If I can't, then it has low SEO value. That feature is very important for me.
952.
▲
by
bravura
14y ago
I am not sure in the sense of: If I were dropped on a desert island, I could derive a water-tight proof of this result from scratch. I am confident , though, based upon my reading of secondary sources written by people that I trust. From
953.
▲
by
bravura
14y ago
Since I see some misunderstanding about deep learning, let me explain the fundamental idea: It's about reusing intermediate work. The intuition is let's say I told you to write a complicated computer program. Let's say I told you that you c
954.
▲
by
bravura
14y ago
SVMs are good if you want high accuracy without much fiddling and don't have many training examples. It is pretty simple to get off-the-shelf results from SVMs. However, SVM training is quadratic in the number of examples, and you have to g
955.
▲
NIPS 2012 (top ML conference) papers available online
(books.nips.cc)
5 points
by
bravura
14y ago
|
0 comments
956.
▲
by
bravura
14y ago
Serious question: I know he's the Woz, but is there any other reason that we care about his opinion? He hasn't been involved with Apple since 1987. Has Woz ever demonstrated any expertise in design or UX? [edit: For anyone that has accused
957.
▲
Ray Kurzweil on the future workforce
(washingtonpost.com)
1 points
by
bravura
14y ago
|
0 comments
958.
▲
by
bravura
14y ago
Well done, Matt. We haven't hung out in a while. Let's get together when I come to SF in January.
959.
▲
by
bravura
14y ago
I believe in using a black background for talk slides. Example: http://new.livestream.com/gigaom/gigaomroadmap/videos/579403... I would prefer dark gray from an aesthetic perspective. But the reason I choose black is simple: Often, the LC
960.
▲
by
bravura
14y ago
Question: What service can I use to find inexpensive flights with flexible travel dates (+-N days) and in-flight wifi? kayak, my typical goto source for finding tickets, allows me to use flexible dates to minimize my travel costs. I then bu
More ›