Search results

yup. you need to have some pretty strong infraestructure just to run the ML systems... and you also need to have a good set of tools to analyze the results, and use that analysis to ask the right questions to the AI... i've been working for the last year full time at setting up such a set...

Rest can't do low latency, but neither can Python... For low latency, he would need to look at a faster, compiled language like java, C/C++, Fortran... a FIX port with Python inside, would be a pretty slow FIX port... if C++ is an option, then Rithmic's C++ API would be a good idea...

You are right, I thought they had one. My mistake.

IB has a RESTful API, it's pretty easy to implement a Python script to talk to it.

I usually go with OpenSource, but that also includes my OS (Fedora <3) so I'm not familiar with Windows tools. Found a good article on partition magic and it's alternatives though. :) en.kioskea.net/faq/1391-free-alternative-to-partition-magic

Sure. You can easily google a step by step guide to do it.

you can install your personal files and folders in one partition and the OS in another, making it easy to format the OS if the need arises without touching your files On UNIX-family OS' you can also share the /home and /var folders across several systems, this is very useful for distro-hopping...

it's the only job I know where friday sucks and monday is awesome. :)

In my old job I had to clean up when a things went wrong for a trader and he was left with a large position on a really thin stock... so watching your video brought back some pretty fun memories. thnx. happy trading.

Oh ok. :) I was really scratching my head thinking that you were just not seeing the liquidity.

any particular reason for now looking at the whole book (just Arca Book and NBBO)?

Written for DOS?? :O maybe this will come in handy... http://www.ccs64.com/

real market data doesn't fit nicely into a straight line, it breaks the line and gets back on top, and then looks choppy for a while before picking up the trend again... the signal is buried in a lot of noise... and you have to sort through that noise... you need to clearly identify the signal...

cool :)

Are you sure the increase is lineal? (double the data=double the time) and not more? (an easier test is to reduce the sample to say, 100 stocks and see if the decrease is lineal or faster) ... ML stands for machine learning...

6 hours? wow.... that's a long time for a few million records... what kind of ML are you running on that data? is that running on a single core, or parallel?

1) cool. R may not be the strongest system, but it makes it pretty easy to get everything up and running. 2) if your dataset is small, then working with dataframes which are kept in RAM is ok; but once the data gets big it starts to slow things down. A database holds the data in the...

Nice trade. Few questions: How come you use only ARB instead of the full book... And why ARB instead of EDGX of Philly where it prints a lot of the volume? Any reason for not using icebergs or dark orders?

Your on the right idea. Back in 2004 I worked booking horse races, and we had a handful of wiseguys who would use stats software to find horses that were paying too much, these were players who had tight bet limits around the books because everyone in the business knew they were they kind of...

Hinton did some research on using NLP on SEC filings and autoencoders This could be an interesting avenue for research since there's plenty of free data through Edgar.