Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
deathflute
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
1.
▲
by
deathflute
1y ago
Lots of debate here about the best model. The best model is the one which creates the most value for you —- this typically is a function of your skill in using the model for tasks that matter to you. Always was. Always will be.
2.
▲
Deep learning networks are approximately kernel machines
(arxiv.org)
2 points
by
deathflute
6y ago
|
1 comments
3.
▲
by
deathflute
6y ago
interview with Pedro: https://news.cs.washington.edu/2020/12/02/uncovering-secrets...
4.
▲
by
deathflute
10y ago
Quantitative trading, hedge fund, New York, NY, ONSITE I run a systematic quant trading group that trades globally. We are research driven and are working on solving tough problems at the intersection of math, statistics, and computer scien
5.
▲
by
deathflute
11y ago
Well said. If you have or build an "edge" over time the distribution of your outcomes will have a positive mean. Along the way you may see values on the tails which are random spikes.
6.
▲
by
deathflute
11y ago
Quantitative trading, hedge fund, New York, NY, ONSITE I am starting a new quantitative trading group that will systematically trade global Futures, FX, and Equities. We are research driven and are working on solving some really tough probl
7.
▲
by
deathflute
11y ago
Agree about the cost. However, I would think that a wider adoption would eventually bring the cost down and perhaps even spawn a bunch of related open-source projects. As far as readability is concerned, q(KDB+) is far more readable than k(
8.
▲
by
deathflute
11y ago
Well, all you need to do is to look at KDB+ from kx. All the other products that you mention are for children ;)
9.
▲
by
deathflute
12y ago
Can somebody comment on the advantages of using this over zeromq/nanomsg? Thanks!
10.
▲
by
deathflute
12y ago
We used them recently in nyc and user experience was great. However once you do the math, a 10 - 15% markup on most products in addition to delivery fees, tips, (or prime) is simply too much if you use them on a regular basis. The annualize
11.
▲
by
deathflute
12y ago
I did not realize that the email is not publicly displayed. I have put in the about section now. hn At machine.imap.cc
12.
▲
by
deathflute
12y ago
You can find my email in my profile by clicking on my username.
13.
▲
by
deathflute
12y ago
Hey, I wanted to send you a message but your profile has no email. Can you send me a message or put an email in your profile? Thanks.
14.
▲
by
deathflute
12y ago
I have a travel startup that is meant to solve the planning problem in a different way, but I agree with Gary. We painfully learned that - a) We created an elegant solution to free. That is always a bad idea b) User acquisition is extremely
15.
▲
by
deathflute
12y ago
Excellent analysis. I am sick of seeing these Nanex posts(rants); the lack a basic familiarity of market microstructure mechanisms. The market maker is not sitting there to let you run him over and thank you for it. As a price taker, the tr
16.
▲
by
deathflute
12y ago
Thanks. That is sort of like going to war with knives amongst machine guns, but what do I know.
17.
▲
by
deathflute
12y ago
A question for DDG or anyone who might know: How does DDG plan to monetize this without storing any data?
18.
▲
by
deathflute
12y ago
Well, the reality is that the top ~5% of hedge funds generate all the returns and have real alpha. The rest are beta + noise - transaction costs. Not dissimilar to the way venture returns are distributed. Only the top startups manage to bec
19.
▲
by
deathflute
12y ago
This would actually make a terrific replacement for something like redis when you need a more structured schema. The q language is very powerful, and expressive - interesting mix of lisp and APL. You can do really powerful analytics without
20.
▲
32 bit version of KDB+ is now free for commercial use
(kx.com)
68 points
by
deathflute
12y ago
|
32 comments
21.
▲
by
deathflute
12y ago
Does anyone know if these scores have been predictive of success in the past?
22.
▲
by
deathflute
13y ago
This is really great advice. One other thing that has helped me immensely is getting a dog! It automatically forces you to take breaks.
23.
▲
by
deathflute
13y ago
"But in fact the main thing driving Airbnb's growth is not price, but that guests are looking for authentic experiences" Do you think that as the services that Airbnb provides gets more standardized, the Airbnb experience wil
24.
▲
by
deathflute
13y ago
Let me pile on the offer. Along with drinks, I can help you analyze your data or advice you on your technology stack.
25.
▲
by
deathflute
13y ago
Thanks for the great advice. We would love to have you look at our application ;)
26.
▲
by
deathflute
13y ago
Thanks. Good advice on getting intros.
27.
▲
by
deathflute
13y ago
Couple of questions for the original poster - 1) Were you already profitable or had revenues when you started fundraising? 2) How did you manage to get intros to Angels?
28.
▲
by
deathflute
13y ago
I see. Any pointers to a cheap payroll processing company?
29.
▲
by
deathflute
13y ago
A naive question to founders who have taken such seed funding before - can you directly take money out of a investment like this to pay for your living costs or you have to go through the headache of establishing payroll and a small salary
30.
▲
by
deathflute
13y ago
Thanks for the clarification. Thats what I thought. So convertible note is much better for the entrepreneur. It retains the optionality of doing well and consequently giving out less equity.
More ›