Re: [Quantlib-dev] Reg. Contributing

[urun d. <uru...@gm...>]

Dear All;

**
>
>
>
> Hi Kim,
>
>> **
>>
>> Hi urun,
>>
>> Am 20.07.2011 22:29, schrieb urun dogan:
>>
>> Hi Luigi,
>> I am a post-doc doctoral researcher in Germany. My main researc focus is
>> machine learning/artificial intelligence techniques. I am a develor of Shark
>> machine learning library.
>>
>>  Just my 2 cent.
>> Since you are involved in the development of the shark machine you
>> definitely know how to use machine learning to classify and to predict
>> observations. If this is the case you should try to get involved into
>> designing and implementing (alpha-) strategies for trading and backtesting.
>>
>
> I have significant amount of knowledge on classification, regression and
> model selection techniques. I made a small research on " (alpha-) strategies
> for trading and backtesting" . They are quite interesting. Is there a group
> implementin these techniques in quant-lib?
>
> Difficult to say. As far as i know there is nobody planing to implement a
> backtesting framework in quantlib.
>
>  If yes where can I found contact details of members?
>
>  If no, should I start by my own?
>
> Hmmm, i remember seeing a webinar from mathwork showing how to use a neural
> network for risk controlling energy trading.
> Perhaps you can try to implement a neural network in ql and use this tool
> to forecast energy prices.
> Just a suggestion.
>

I have implemented several feedforward neural network methods, recurrent
neural network methods, support vector machines and so on. I think this line
of implmentation can be beneficial to some people.
I am reading some documents/papers about algorithmic trading and I think
these methods can be also used in algorithmic trading. To put such kind of
methods in quantlib is very interesting for me.
I have three questions. First one is: Where can I find some data like energy
trading data or finance data? The second one is ehich methods are the state
of art nethods for this kind of tasks? The third one is
which methods have more priority than others? I mean what are the industry
requirement? Does quantitative developers/analysers/traders need very fast
and accurate methods, e.g. Neural networks, conventional
linear regression methods or accurate but slow methods e.g SVMs, Gaussian
Process?

Best regards
Urun


>>
>>  This library is implemented by C++ . I am really interested in
>> contributing to quant lib because I find find finsnce chalenging and
>> interesting. I am open to implement some machine learning techniques to
>> quant lib also it is absolutelly ok for me to implement other things. Is
>> there any todo list? Are there any ideas for starting?
>>
>> Thanks a lot for your help.
>>
>> Best regards
>> Ueruen Dogan
>> On 18 Jul 2011 13:40, "Luigi Ballabio" <lui...@gm...> wrote:
>>
>>
>> ------------------------------------------------------------------------------
>> 10 Tips for Better Web Security
>> Learn 10 ways to better secure your business today. Topics covered include:
>> Web security, SSL, hacker attacks & Denial of Service (DoS), private keys,
>> security Microsoft Exchange, secure Instant Messaging, and much more.http://www.accelacomm.com/jaw/sfnl/114/51426210/
>>
>>
>> _______________________________________________
>> QuantLib-dev mailing lis...@li...://lists.sourceforge.net/lists/listinfo/quantlib-dev
>>
>>
>>
>
>