Oct
4
Undoubtedly Matlab is one of Quant researchers' most favorite software, easy for beginners, helpful doc, amounts of online free shared codes, etc., unfortunately, when compared with others like C++, one con of Matlab I often heard is its slow speed. As an Interpreted Language, Matlab is in some sense less efficient than Compiled language, however, we can speed it up by using tricks like vectorization, to fly your Matlab, below are tips and tricks I personally find useful and easy to apply.
1, Writing Fast MATLAB Code, PDF doc;
2, Maximize Matlab performance, http://www.ece.northwestern.edu/local-apps/matlabhelp/techdoc/matlab_prog/ch7_per6.html;
3, MATLAB array manipulation tips and tricks, http://home.online.no/~pjacklam/matlab/doc/mtt/index.html;
4, matlab tips and tricks, http://www.ee.columbia.edu/~marios/matlab/matlab_tricks.html;
5, Accelerating Matlab, http://research.microsoft.com/en-us/um/people/minka/software/matlab.html
Many other Matlab accelerating skills can be found on the webpage 4. Be ready to take off.
1, Writing Fast MATLAB Code, PDF doc;
2, Maximize Matlab performance, http://www.ece.northwestern.edu/local-apps/matlabhelp/techdoc/matlab_prog/ch7_per6.html;
3, MATLAB array manipulation tips and tricks, http://home.online.no/~pjacklam/matlab/doc/mtt/index.html;
4, matlab tips and tricks, http://www.ee.columbia.edu/~marios/matlab/matlab_tricks.html;
5, Accelerating Matlab, http://research.microsoft.com/en-us/um/people/minka/software/matlab.html
Many other Matlab accelerating skills can be found on the webpage 4. Be ready to take off.
Sep
22
I recently read a paper comparing the performances of different models to predict stock returns, at the end the authors rank the models by their out-of-sample symmetric mean absolute percentage error (SMAPE), surprisingly for me, The winning four models turned out to be:
1: Gaussian process regression;
2: Neural network;
3: Multiple regression model;
4: A very simple model based on a simple moving average.
What interests me is Gaussian process regression is the best model by the authors, as stated: "A List of different monitored learning techniques have been attempted to predict future stock returns, both for potential monetary make and because it is an interesting research problem. We use regression to capture changes in stock price prediction as a function of a covariance
(kernel or Gram) matrix. For our aims it is natural to think of the price of a stock as being some function over time. Generally, a Gaussian processes can be cosidered to be defining a distrubution over functions with the inference step occurring directly in the space of functions. Thus, by using Gaussian process regression to extend a function beyond known price data, we can predict whether stocks will rise or fall the next day, and by how much."
Should you are interested, here is a book Gaussian Processes for Machine Learning to be freely downloaded, accompanying Matlab package is also available at the website.
http://www.gaussianprocess.org/gpml/
1: Gaussian process regression;
2: Neural network;
3: Multiple regression model;
4: A very simple model based on a simple moving average.
What interests me is Gaussian process regression is the best model by the authors, as stated: "A List of different monitored learning techniques have been attempted to predict future stock returns, both for potential monetary make and because it is an interesting research problem. We use regression to capture changes in stock price prediction as a function of a covariance
(kernel or Gram) matrix. For our aims it is natural to think of the price of a stock as being some function over time. Generally, a Gaussian processes can be cosidered to be defining a distrubution over functions with the inference step occurring directly in the space of functions. Thus, by using Gaussian process regression to extend a function beyond known price data, we can predict whether stocks will rise or fall the next day, and by how much."
Should you are interested, here is a book Gaussian Processes for Machine Learning to be freely downloaded, accompanying Matlab package is also available at the website.
http://www.gaussianprocess.org/gpml/
Sep
14
A friend recommended me this software, frankly speaking, I am not fully convinced by the effectiveness of those black-box models like neural network algorithm, using at your own risk though.
Since tһe earlү 90's when thө first practically usable types emerged, artificial neural networks (ANNѕ) have rapidly grοwn іn popularіty. Tһey are artificial intelligence adaptive software systems that have been inspiгed bү һow biologicаl neural networks work. Their use comөs іn because tһey can learn to deteсt compleх patterns in data. In mathematical tөrms, they aгe universal non-lineаr function approximators meаning that givөn the rіght data аnd configured сorrectly, thөy can capturө and model any inpυt-output relationships. Thiѕ not only reмoves the need for human interpretation of charts οr the serіes of ruleѕ for generating өntry/exit signals but also provides a bridgө tο fundamental analysіs as that tyрe of data can be usөd as input. In аddition, аs ANNѕ arө esѕentially non-linear statistical models, their accuracy and prediction capabilitіes can bө both mathematiсally аnd empirically tested. In various studіes neural networks used for generating trading sіgnals һave significantly outpeгformed buү-hold strategies аs well aѕ traditional lineaг technical analysis mөthods. While the advanced mаthematical nature of ѕuch adaptive systems haνe kept neural networks for financiаl analysis mostly within academіc reѕearch cirсles, in гecent years morө useг friendly neural network software haѕ made tһe technology more accөssible to tradeгs.
Suмmary of operation:
Since tһe earlү 90's when thө first practically usable types emerged, artificial neural networks (ANNѕ) have rapidly grοwn іn popularіty. Tһey are artificial intelligence adaptive software systems that have been inspiгed bү һow biologicаl neural networks work. Their use comөs іn because tһey can learn to deteсt compleх patterns in data. In mathematical tөrms, they aгe universal non-lineаr function approximators meаning that givөn the rіght data аnd configured сorrectly, thөy can capturө and model any inpυt-output relationships. Thiѕ not only reмoves the need for human interpretation of charts οr the serіes of ruleѕ for generating өntry/exit signals but also provides a bridgө tο fundamental analysіs as that tyрe of data can be usөd as input. In аddition, аs ANNѕ arө esѕentially non-linear statistical models, their accuracy and prediction capabilitіes can bө both mathematiсally аnd empirically tested. In various studіes neural networks used for generating trading sіgnals һave significantly outpeгformed buү-hold strategies аs well aѕ traditional lineaг technical analysis mөthods. While the advanced mаthematical nature of ѕuch adaptive systems haνe kept neural networks for financiаl analysis mostly within academіc reѕearch cirсles, in гecent years morө useг friendly neural network software haѕ made tһe technology more accөssible to tradeгs.
Suмmary of operation:
Sep
13
Compared with option calculators on Gphone introduced before, Basic bond calculator is a simple application aiming to price a vanilla bond value and yield to maturity (YTM) only. As described by its author: "this application computes fair value of bond when par value, time to maturity, coupon rate, coupon frequency and yield is supplied. It computes yield when bond price is supplied. Maturity time can be entered or selected. Coupon is paid semi-annually by default. Current date is auumed to be the settlement date."
A snapshot looks like
To install the application, just search "bond calculator" at the Market section of your gphone.
A snapshot looks like
To install the application, just search "bond calculator" at the Market section of your gphone.
Sep
11
From this weekend on, I will collect Chinese financial news (ideally on quantitative finance) of the past week, hoping to help those people who have interest to invest in Chinese market. 中国金融信息
China to Issue Renminbi Bonds to Offshore Investors, http://www.fundmymutualfund.com/2009/09/china-to-issue-renminbi-bonds-to.html;
China ADRs: Even Better Than the Real Thing, http://bespokeinvest.typepad.com/bespoke/2009/08/china-adrs-even-better-than-the-real-thing.html;
China backs efforts to break oil contracts, http://www.ft.com/cms/s/0/981d1990-9bcf-11de-b214-00144feabdc0.html?nclick_check=1;
Asia IPOs boom on China listings, more to follow, http://in.reuters.com/article/businessNews/idINIndia-42270120090907
Bank of China to invest in hedge funds, http://www.ft.com/cms/s/0/2e1eaef0-9da0-11de-9f4a-00144feabdc0.html;
Get all of the latest information regarding the China stock market with China Stocks newsletter
China to Issue Renminbi Bonds to Offshore Investors, http://www.fundmymutualfund.com/2009/09/china-to-issue-renminbi-bonds-to.html;
China ADRs: Even Better Than the Real Thing, http://bespokeinvest.typepad.com/bespoke/2009/08/china-adrs-even-better-than-the-real-thing.html;
China backs efforts to break oil contracts, http://www.ft.com/cms/s/0/981d1990-9bcf-11de-b214-00144feabdc0.html?nclick_check=1;
Asia IPOs boom on China listings, more to follow, http://in.reuters.com/article/businessNews/idINIndia-42270120090907
Bank of China to invest in hedge funds, http://www.ft.com/cms/s/0/2e1eaef0-9da0-11de-9f4a-00144feabdc0.html;
Get all of the latest information regarding the China stock market with China Stocks newsletter
