Possible future works
Locked
 
|
Hi All, Having learned QuantLib for a while, I’m very impressed by the great design of the software architecture, and I would love to contribute to this project. I’m a software engineer from general signal processing background, but I don’t really have much experience on quantitative finance. I have read the developers mail list archive but could not find any topic related to possible future works that a developer like me can contribute to. So I would be really appreciated if someone can let me know what is currently missing and will be nice to have in future QuantLib. A list of these possible future works will surely benefit the development of QuantLib. To start with, do you think a general Expectation-Maximization algorithm could be added into the future to-do-list? Thanks and regards, James ------------------------------------------------------------------------------ Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk _______________________________________________ QuantLib-dev mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-dev |
|
|
It seems that EM algo can be found here:
http://code.opencv.org/projects/opencv/repository/revisions/master/entry/modules/ml/src/em.cpp#L126 On Thu, December 18, 2014 3:09 pm, James wrote: > Hi All, > > > > Having learned QuantLib for a while, I'm very impressed by the great > design > of the software architecture, and I would love to contribute to this > project. I'm a software engineer from general signal processing > background, > but I don't really have much experience on quantitative finance. I have > read the developers mail list archive but could not find any topic related > to possible future works that a developer like me can contribute to. So I > would be really appreciated if someone can let me know what is currently > missing and will be nice to have in future QuantLib. A list of these > possible future works will surely benefit the development of QuantLib. > > > > To start with, do you think a general Expectation-Maximization algorithm > could be added into the future to-do-list? > > > > Thanks and regards, > > James > > ------------------------------------------------------------------------------ > Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server > from Actuate! Instantly Supercharge Your Business Reports and Dashboards > with Interactivity, Sharing, Native Excel Exports, App Integration & more > Get technology previously reserved for billion-dollar corporations, FREE > http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk_______________________________________________ > QuantLib-dev mailing list > [hidden email] > https://lists.sourceforge.net/lists/listinfo/quantlib-dev > -- Xiang Ni, Department of Mathematics California Institute of Technology http://math.caltech.edu/people/xiang-homepage.html ------------------------------------------------------------------------------ Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk _______________________________________________ QuantLib-dev mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-dev |
|
|
Hi Xiang Ni,
Thanks for your reply. I'm not surprised that there is a implementation already available. Actually, you can probably find available implementations for most algorithms that we are aware of. But whether the code we find fits the framework (QuantLib) is a question. For instance, the data structures used throughout the algorithm may need to be converted, and this task alone may not be trivial... Whether an algorithm has already been implemented or not is a bit off the topic. I think a more important question should be - is this algorithm really needed for QuantLib? Regards, James -----Original Message----- From: Xiang Ni [mailto:[hidden email]] Sent: Friday, December 19, 2014 12:16 AM To: James Cc: [hidden email] Subject: Re: [Quantlib-dev] Possible future works It seems that EM algo can be found here: http://code.opencv.org/projects/opencv/repository/revisions/master/entry/modules/ml/src/em.cpp#L126 On Thu, December 18, 2014 3:09 pm, James wrote: > Hi All, > > > > Having learned QuantLib for a while, I'm very impressed by the great > design of the software architecture, and I would love to contribute to > this project. I'm a software engineer from general signal processing > background, but I don't really have much experience on quantitative > finance. I have read the developers mail list archive but could not > find any topic related to possible future works that a developer like > me can contribute to. So I would be really appreciated if someone can > let me know what is currently missing and will be nice to have in > future QuantLib. A list of these possible future works will surely > benefit the development of QuantLib. > > > > To start with, do you think a general Expectation-Maximization > algorithm could be added into the future to-do-list? > > > > Thanks and regards, > > James > > ---------------------------------------------------------------------- > -------- Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT > Server from Actuate! Instantly Supercharge Your Business Reports and > Dashboards with Interactivity, Sharing, Native Excel Exports, App > Integration & more Get technology previously reserved for > billion-dollar corporations, FREE > http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg. > clktrk_______________________________________________ > QuantLib-dev mailing list > [hidden email] > https://lists.sourceforge.net/lists/listinfo/quantlib-dev > -- Xiang Ni, Department of Mathematics California Institute of Technology http://math.caltech.edu/people/xiang-homepage.html ------------------------------------------------------------------------------ Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk _______________________________________________ QuantLib-dev mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-dev |
|
|
Hello, do you have a financial use case for the algorithm? Luigi On Sat, Dec 20, 2014 at 11:58 PM, James <[hidden email]> wrote: Hi Xiang Ni, ------------------------------------------------------------------------------ Dive into the World of Parallel Programming! The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net _______________________________________________ QuantLib-dev mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-dev |
|
|
Hi Luigi, Yes, there are a few financial use cases for the Expectation-Maximization (EM) algorithm. For instance, it can be used to train models of stock market interactions. If we use a model to capture the interactions between the stock markets, EM algorithm can be used to iteratively improve the model until some parameter closely models the data, or until the algorithm times out. Here is a link that illustrate the above example: http://www.tryfsharp.org/Learn/financial-computing#training-the-model I’m not from finance background and I’m stilling learning QuantLib, so my proposal may not really fit this library. But please let me know if this could be useful and I will start working on it. Thanks and regards, James From: Luigi Ballabio [mailto:[hidden email]] Hello, do you have a financial use case for the algorithm? Luigi On Sat, Dec 20, 2014 at 11:58 PM, James <[hidden email]> wrote:
-- ------------------------------------------------------------------------------ Dive into the World of Parallel Programming! The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net _______________________________________________ QuantLib-dev mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-dev |
|
|
In reply to this post by james-2
Hi Xiang Ni,
I'm trying to use quantlib for some very basic algo trading with bitcoin futures. My wife is also trading HK options, and I'm trying to write some very basic option pricing systems for her. My general impression is that it would be better to keep quantlib a pure derivatives pricing library, and that anything involving machine learning or algo trading, would be better off in a library that is separate from quantlib. The way that I'm structuring my code is that I'm using the python-SWIG interface to quantlib, and using quantlib only for pricing. Everything else I'm keeping outside of quantlib, and interface it with quantlib. So it would make sense to me to write your code as a separate library with python interfaces and then use it to glue it to quantlib. I have a docker image (joequant/bitstation-prod). that has a installed python installation with quantlib and a lot of other goodies installed. ------------------------------------------------------------------------------ Dive into the World of Parallel Programming! The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net _______________________________________________ QuantLib-dev mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-dev |
|
|
In reply to this post by james-2
FYI, I worked at a major investment bank until a year ago when I
started my own company, and the derivatives trading pricing code had only one feature that quantlib doesn't have, and that's the ability to dynamically load in parts of the code. The issue is that because all of the derivatives code is in one library, it becomes impossible to load in the entire pricing library. So there is a slick infrastructure in which the derivatives library is modularized into dozens of small dynamic modules. When a system needs to run a c++ class, the system tracks which module it is in and it autoloads that module. This was also a bit tricky since it had to run on both Windows and linux systems. It would be interesting to see if this could be implemented in QuantLib. The autoloading mechanism was written by one person who took a few months to do it. Personally, I don't think it would be worth while to write an autoload mechanism from scratch, but I'd be interested in knowing if there was some off the shelf system that could be adapted for quantlib. Something that would be more low hanging fruit which would be easier would be to just generate several shared libraries instead of one big shared library or to have a mechanism by which someone can add a class to QuantLib, put it into a shared library and be able to work on that without having to do a complete recompile/relink of the library. ------------------------------------------------------------------------------ Dive into the World of Parallel Programming! The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net _______________________________________________ QuantLib-dev mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-dev |
«
Return to quantlib-dev
|
1 view|%1 views
| Free forum by Nabble | Edit this page |