Thanks for the help Sina and Black:This is just my closing blurb on what I found and my direction going forward. I reviewed all the articles and websites Sina sent me and many others. I have decided to do my AI work in Vb.net, or if necessary C#. My Reasoning: There are plenty of code examples and libraries for them, I am very comfortable with the languages and IDE and, finally, the whole subject of Neural Networks is going to take massive research and experimentation on my part. This way I can eliminate the learning curve of LISP, perhaps more than one dialect and testing various compilers or IDE products for accessibility, execution speed and features. In addition the diferent third-party products and libraries all seem to use diferent forms of Back-Propigation, Genetic and Linear processing as well as other algorithms. This is the same for VS so is a wash in my decision. So, I am already set up to start programming a Neural Network from the ground up with some very simple examples which will help me learn the subject matter without first setting up and learning a LISP Environment. I can concentrate on the proper use of the algorithms, Data acquisition, Model Testing and various results from various forms of the basic NN types. That is, Predictive, Category and others I forget off-hand. So thanks for the help in making my decision on how to break into this field and I'm off to look for some more sample projects, libraries to pick one to begin with.
Rick USA----- Original Message ----- From: "Sina Bahram" <sbahram@xxxxxxxxx>
To: <programmingblind@xxxxxxxxxxxxx> Sent: Wednesday, October 15, 2008 6:39 PM Subject: RE: Neural Network Programming Overview Question
Forward them along if you don't mind. Take care, Sina -----Original Message----- From: programmingblind-bounce@xxxxxxxxxxxxx [mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of black ares Sent: Wednesday, October 15, 2008 6:15 PM To: programmingblind@xxxxxxxxxxxxx Subject: Re: Neural Network Programming Overview Question I think I have some where some Ai books and genetic algorithms. Please e-mail offlist and I will try to answer with some books.----- Original Message ----- From: "Sina Bahram" <sbahram@xxxxxxxxx>To: <programmingblind@xxxxxxxxxxxxx> Sent: Wednesday, October 15, 2008 11:01 PM Subject: RE: Neural Network Programming Overview QuestionThere are probably millions of lines of lisp you can pull from for libraries and things of that nature for neural networks of varying types and flavors.Please note that's equivalent, literally, to hundreds of millions of linesof code of C# or other lesser languages, *smile*. Here are some random, and not so random links below. An interesting guy who has done research related to what you're asking about. http://www.idsia.ch/~juergen/ A common strategy, and one of the few that actually has a chance in my opinion, is evolutionary techniques that evolve neural networks. To that end, here you go. Neurocomputing : Neural network construction and training using grammatical evolution:http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6V10-4S1C894-3&_user=10&_rdoc=1&_fmt=&_orig=search&_sort=d&view=c&_version=1&_urlVersion=0&_userid=10&md5=9736cb76398111c4e1306f5d31ec18de Watch out if links wrap ... That one's a long one, for example. Using multi-agents to predict the stock market evolution based on fundamentalist analysis and fuzzy-neural networks http://portal.acm.org/citation.cfm?id=1366414 Designing neural networks using genetic algorithms:http://citeseerx.ist.psu.edu/showciting;jsessionid=6F2C4C9E46A26B6323421B3062D5B2F2?cid=65706 Here's a general collection that might be useful: Keith Price Bibliography New Unsorted Entries, and Other Miscellaneous Papers Comparison of artificial intelligence & machine learning algorithms as a predictor of surgical outcomes in benign prostatic hperlasia cases (BPH): http://cat.inist.fr/?aModele=afficheN&cpsidt=14046882 An Analysis of Factors Directing the Admission Process of Artificial Intelligence Technologies http://dbpubs.stanford.edu:8090/pub/1995-16 Note: I think that last one has the full text available. Stock Market Modeling Using Genetic Programming Ensembles http://www.springerlink.com/content/w061315720425n76/ Connectionism http://www.linuxselfhelp.com/HOWTO/AI-Alife-HOWTO-3.html Now for some code related links: Common-Lisp.net http://common-lisp.net/ C liki wiki: http://www.cliki.net/index The Common Lisp Open Code Collection (CLOCC) http://clocc.sourceforge.net/The common lisp hyperspec, this is most likely the most important link outof the whole bunch: http://www.lispworks.com/documentation/HyperSpec/index.html Tour de Lisphttp://tourdelisp.blogspot.com/2008/01/common-lisp-libraries-victims-of-drive.html Read the comments on that site, too. Anyways, hope that helps. Take care, Sina ________________________________ From: programmingblind-bounce@xxxxxxxxxxxxx [mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of Ricks Place Sent: Wednesday, October 15, 2008 2:53 PM To: programmingblind@xxxxxxxxxxxxx Subject: Neural Network Programming Overview Question Hi Guys: See if this sounds right: A neural network may be any of several types.(models) MLP FeedForward, BiDirectional models and many other Dynamic variations. Inputs can be Filtered, weighted and tuned using Back-Propagation, Genetics, Fuzzy Logic and Chaos Methods which seem to be the ones best suited to Stock Market and Asset Allocation based on Technicals and fundementals. LISP is the programming language of choice for most programmers in the AI field but is not supported in Visual Studio. If I hand code I would have to code all of the models of a given network if I wanted to evaluate each model's effectivness in prediction of optimum Asset Allocation or Market Timing. There may, or not, be modules, dlls? out there for Back-Propagation, Genetics and Fuzzy Logic I could use in a home grown LISP application. There might be some LISP based development environments for developing Neural Networks available with embedded heavy math functions already set up to apply against inputs for a LISP application. Otherwise I would have to design and code them. Do I have a generalized overview of the situation? I am trying to decide on developing in LISP using just a LISP environment, a specialized environment, a plug in for developing NNs in Visual Studio, a Stand Alone system like NeuroShell or something else like just subscribing to a service for investing using the NeuroShell models. Thanks Guys - this is really big time and I think I am just out of my league but want to see if I have the general ideas before making any initial decision costing allot of time, money or both. Rick USA __________ View the list's information and change your settings at //www.freelists.org/list/programmingblind__________ View the list's information and change your settings at //www.freelists.org/list/programmingblind __________ View the list's information and change your settings at //www.freelists.org/list/programmingblind
__________View the list's information and change your settings at //www.freelists.org/list/programmingblind