COMPGROUPS.NET | Post Question | About | Groups | Contact | Register | Login



comp soft-sys matlab (18537)

comp os linux advocacy (5162)

comp soft-sys sas (3808)

comp soft-sys math mathematica (2920)

comp lang python (2910)

comp lang ruby (2557)

comp text tex (1734)

comp sys mac advocacy (1660)

comp lang c++ (1422)

comp lang javascript (1223)

comp lang c (1197)

comp lang perl misc (1192)

comp lang java programmer (927)

comp dsp (911)

comp sys mac system (893)

comp lang tcl (873)

comp databases ms-access (764)

comp arch fpga (751)

comp lang idl-pvwave (727)

comp software shareware announce (704)

comp lang php (700)

comp lang lisp (684)

comp sys sun announce (659)

comp arch embedded (626)

comp sys mac apps (613)

comp sys ibm ps2 hardware (545)

comp lang fortran (525)

comp lang c++ moderated (524)

comp unix shell (523)

comp os plan9 (514)

comp os linux misc (513)

comp unix solaris (501)

comp databases oracle server (497)

comp sys ibm pc games war-historical (485)

comp lang python announce (483)

comp os ms-windows programmer win32 (477)

comp databases mysql (412)

comp sys cbm (408)

comp dcom sys cisco (408)

comp arch (308)

comp sys ibm pc games action (301)

comp databases informix (274)

comp lang clipper visual-objects (267)

comp os vms (263)

comp lang perl announce (238)

comp lang xharbour (229)

comp graphics apps gnuplot (224)

comp realtime (216)

comp soft-sys ace (215)

comp lang ada (213)

comp protocols nfs (207)

comp protocols snmp (204)

comp parallel (202)

comp sys acorn announce (196)

comp protocols tcp-ip domains (193)

comp sys ibm pc hardware systems (191)

comp unix internals (187)

comp infosystems www browsers misc (184)

comp misc (182)

comp publish cdrom software (182)

comp dcom cabling (174)

comp mail mime (169)

comp std announce (169)

comp ai fuzzy (168)

comp ai edu (166)

comp soft-sys stat spss (156)

comp protocols kerberos (155)

comp databases filemaker (155)

comp sys apple2 (153)

comp admin policy (151)

comp databases postgresql committers (150)

comp databases ibm-db2 (149)

comp databases ms-sqlserver (131)

comp protocols dicom (129)

comp unix programmer (128)

comp security misc (123)

comp unix bsd freebsd misc (120)

comp graphics visualization (117)

comp cad cadence (117)

comp databases olap (116)

comp apps spreadsheets (114)

comp databases postgresql general (114)

comp lang vhdl (113)

comp lang java help (113)

comp ai neural-nets (112)

comp ai (111)

comp ai games (110)

comp dcom telecom (109)

comp client-server (108)

comp os linux announce (105)

comp cad microstation (105)

comp databases oracle marketplace (105)

comp soft-sys math scilab (102)

comp databases progress (102)

comp os linux networking (101)

comp database oracle (100)

comp graphics api opengl (99)

comp ai genetic (98)

comp protocols dns bind (97)

comp compilers (97)


comp.ai.neural-nets Post New
Items(93) /2 Next >> Last >|
Subject Posted Replies From
how to build a network - bioinformatics

hi, how to build a network, if the input I have 20 amino acid sequence, and only one output neuron, which can only be in a position to 0 or 1? I have a learning network of 20 sequences that always gives 1. then analyze the 1000 sequence and in response I get 0 or 1. thank you rafal

3/30/2010 8:59:16 PM 1 rafal pawlak <rafal.b.paw...@gmail.com>
Multiple Back-Propagation Software - Version 2.2 available

Version 2.2 of Multiple Back-Propagation (MBP) is available for download either at http://dit.ipg.pt/MBP or alternatively at http://sourceforge.net/projects/mbp/ This version supports training networks with data containing missing values (data must be on a CSV file). A value is considered as being absent (missing) if its value is not present or if it is equal to a question mark (?). To deal with missing values MBP does not use imputation or any other pre-processing technique, but it rather creates a network with =93selective inputs=94, which can handle directly missing values. Selec

3/30/2010 11:22:32 AM 0 Noel Lopes <noel.lopes....@gmail.com>
hai

i m rohini now i m doing my ph.d in neural newtworks i m learning from basics now i dont no how to find target values in backpropagation model help to know this with any examples(values) And i want to solve this also A B C D=[a1 b1 c1 d1] a2 b2 c2 d2 a3 b3 c3 d3 a4 b4 c4 d4

3/29/2010 10:19:45 AM 0 naveena lakshmi <anapa.nave...@gmail.com>
CFP: International Symposium on Open Source Intelligence & Web Mining 2010

International Symposium on Open Source Intelligence & Web Mining 2010 http://www.osint-wm.org/ After the great success of OSINT-WM 2008 and OSINT-WM 2009, which were held in conjunction with the 12th International Conference Information Visualization, IV08 in London, UK and 13th International Conference Information Visualization, IV09 in Barcelona, Spain, we look forward to a more successful OSINT-WM 2010 http://osint-wm.org/index.html to be held in Conjunction with the ASONAM 2010. The aim of this symposium is to provide an international forum for researchers, profession

3/27/2010 12:59:10 PM 0 "tansel...@gmail.com" <tansel...@gmail.com>
Technical Analysis Software (TOP 30 - March 2010)

6000 technical analysis programs, indicators, trading systems (64 CDs RAR-archives) 3000 books, courses, seminars (34 DVD RAR-archives) http://stas.ws TOP 30: ======= OmniTrader 2010 RT and VisualTrader 5.0 (all add-ons + courses, seminars) 7 CDs TradeStation 8.7 Build 3085 + OwnData 2.7 + more than 1000 systems 6 CDs MetaStock Pro 11.0 RT for eSignal and more that 200 plug-ins Ocean Combo for TS by Jim Sloman + eBook + video-course MultiCharts 5.5.2723.400 and 6.0 + all add-ons NinjaTrader 6.5 / 7.0 and all add-ons MTPredictor 6.5 EOD + RT Elwave 9.2 KwikPop 7.1 NetPicks 3.

3/27/2010 2:41:50 AM 0 Forex Club <anon6...@yahoo.com>
Best possible classification rate on the PROBEN1 / HEART1 data set

I am trying out some machine learning techniques on the PROBEN1 benchmark, specifically the HEART1 test set. I noticed that several papers report results with about 18-20% test error. Is it possible to achieve better classification rates? Kind Regards Theo

3/23/2010 8:56:40 AM 2 theinze <theo.hei...@googlemail.com>
holographic NN's

Does anyone know of a public domain 'holographic NN' like And Corporation used to sell? I am enthralled with that method. I've tried to reach the original author, Mr. Sutherland, again but have no success now. Thanks,

3/18/2010 12:48:55 AM 0 David <ai...@knology.net>
Some basics

I've studied NN's for many years and even played with a few. However, I cannot make them work in a real experiment because I just can't get my head around translating real-world values into some kind of across-the-board standard values that the NN's use. For example, how would one convert a picture of just a face into a value "between 0 and 1" or "-1 and 1," while keeping relative consistency with translating engine rpm to "0 and 1," and wind speed "between 0 and 1?" Hope this makes sense. Sometimes I wonder if it matters if there IS consistency across the data domains wh

3/18/2010 12:46:39 AM 0 David <ai...@knology.net>
CFP: Intelligent Methods for Protecting Privacy and Confidentiality in Data

http://www.ehealthinformation.ca/CAI/index.asp May 30th, 2010, Ottawa, held in conjunction with Canadian AI 2010. The workshop invites position papers describing original work in theory and applications of intelligent methods to the problem of Data Leak Prevention (DLP). Position papers will be reviewed by the Program Committee members according to their originality, technical merit and clarity of presentation. Each accepted paper will be allocated a maximum of 5 pages in the workshop proceedings. At least one author for each accepted paper is expected to attend the worksho

3/15/2010 4:43:42 PM 0 Marina Sokolova <sokolova.mar...@gmail.com>
EasyNN-plus V13 free update

Coming in the next update. Changes since the half price offer. The update will be free to all users of version 13 including those who bought it in the half price offer. 1. Fixed the crash when right click on empty grid. 2. Disabled Validating Rules when grid is empty. 3. Fixed a looping problem in View > Graph > Column Values after Undo. 4. Fixed the problems with setting and executing a function on a grid range. 5. Added more detail to messages when importing xls files. 6. Added warning message for missing headers when importing xls files. 7. Removed right click Se

3/15/2010 12:33:51 PM 0 Stephen Wolstenholme <st...@tropheus.demon.co.uk>
Einstein's Biggest Blunder - - Wasn't??

Einstein's Biggest Blunder - - Wasn't?? a bit more light about it (click link) http://www.amperefitz.com/einsteins.blunder.htm Enjoy, Fitz

3/14/2010 8:41:28 PM 0 fitz <zeus...@yahoo.com>
Call for Papers: 2010 International Conference on Computer Applications and Industrial Electronics (ICCAIE 2010)

We apologize if you receive multiple copies of this message. Appreciate it if you can disseminate to your peers. Dear Sir/Madam, ======================= CALL FOR PAPER ======================= It is our great pleasure to announce the 2010 International Conference on Computer Applications and Industrial Electronics (ICCAIE 2010) to be held in Kuala Lumpur, MALAYSIA. between the 5th-8th December 2010. The conference will be organized to invite international delegates, to share their latest research findings besides getting to know Malaysia. ICCAIE 2010 will be organized by the

3/14/2010 2:00:00 PM 0 Ihsan <ihsan.yas...@gmail.com>
Neural Network Researchers: Collaborate with over 13,000 researchers for free

Dear Neural Network Researcher, I created MyNetResearch.com to bring about efficient research management and collaboration. It has quickly grown to over 13,000 members from 95 countries. With MyNetResearch.com, you can: - Manage all your research through a modern dashboard - Search for and contact over13,000 other researchers to collaborate with - Share your knowledge and learn from others - Conduct online surveys and citation searches - Improve your research and statistical design using expert tools Join thousands of researchers who create their papers and grant pro

3/11/2010 5:14:40 PM 0 Arin280 <arin...@gmail.com>
Bootstrapping an ANN

Dear neural netters, I have a question regarding the bootstrapping of an ANN. The ANN applied in the context of my study used a back-propagation algorithm (nevprop) that consisted of five input variables, four hidden layers containing five neurons each, and one output layer. My inputs were standardized (0 to 1) to facilitate inclusion of an age variable and weights were initially set using a random seed value. Optimization of the network was undertaken using bootstrapping methods, with 50% of the training data removed from the training set and used to assess the accuracy of the ne

3/10/2010 12:51:11 PM 0 Wes_Roberts <gimps....@gmail.com>
complex function

Dear Colleague, pls help me to plot the complex segmoid function surface with matlab. thanks

3/10/2010 11:04:29 AM 0 blue <sra7...@gmail.com>
1st announcement: WSOM 2011, Workshop on Self-Organizing Maps

8th WORKSHOP ON SELF-ORGANIZING MAPS WSOM 2011, Espoo, Finland Venue: Aalto University School of Science and Technology and Dipoli Conference Center 13-15 June 2011 GENERAL INFORMATION WSOM 2011 will bring together researchers and practitioners in the field of self-organizing systems, with a particular emphasis on the self-organizing maps. It will highlight key advances in these and closely related fields. WSOM 2011 is the eighth conference in a series of bi-annual international conferences started with WSOM'97 in Helsinki. The event will be co-located with the ICANN 2011 c

3/5/2010 1:41:25 PM 0 timoh <timo.honk...@hut.fi>
CFP: ITBI-10 (12-14, Nov. 2010; Nagpur, India)

CALL FOR PAPERS 2nd International conference on IT and Business intelligence (ITBI-10) 12-14, Nov. 2010; Nagpur, India Websites: http://imtnagpur.ac.in/itbi10/

2/27/2010 3:29:16 AM 0 ISNN2010 <isn...@gmail.com>
five year old obituary notice

On Thu, 25 Feb 2010, News Source <anonymous@not-for-mail.invalid> wrote: >Google Groups: How Not to Run an Archive > >http://www.kevinalfredstrom.com/2010/02/google-groups-how-not-to-run-an-archive/ > >by Kevin Alfred Strom > >IT'S BEING acknowledged everywhere - now by Wired magazine and the >SEO Theory newsletter. Google Groups is totally broken. > >http://www.wired.co.uk/news/archive/2009-10/08/the-lost-archives-of-google-groups.aspx > >http://www.seo-theory.com/2008/12/30/how-google-groups-search-has-become-virtually-worthless/ > >It began well. Google acquired the Deja

2/26/2010 5:21:29 PM 0 "Anonymous Remailer (austria)" <mixmas...@remailer.privacy.at>
Call for Papers: The 2010 International Conference of Systems Biology and Bioengineering (ICSBB 2010)

CFP: The 2010 International Conference of Systems Biology and Bioengineering (ICSBB 2010) From: IAENG - International Association of Engineers http://www.iaeng.org/WCE2010/ICSBB2010.html Draft Paper Submission Deadline: 6 March, 2010 Camera-Ready papers & Registration Deadline: 31 March, 2010 WCE 2010: London, U.K., 30 June -2 July, 2010 The conference ICSBB=9210 is held under the World Congress on Engineering 2010. The WCE 2010 is organized by the International Association of Engineers (IAENG), a non-profit international association for the engineers and the computer scientists.

2/23/2010 2:06:02 PM 0 IMECS <imecs_2...@iaeng.org>
2012: It's Only The End Of The World

-----BEGIN PGP SIGNED MESSAGE----- As the pseudochristians and other Antichrist Atheists around the Fallen world scramble to avoid the imminent universal desolation, it's gratifying to watch this international Empire of Antichrist being universally destroyed more of flesh and bone than of walls and stone. And there's absolutely nothing they can do to stop it. Their universal desolation is inevitable unto the culmination of

2/23/2010 3:38:19 AM 1 "Anonymous Remailer (austria)" <mixmas...@remailer.privacy.at>
2012: It's Only The End Of The World

-----BEGIN PGP SIGNED MESSAGE----- As the pseudochristians and other Antichrist Atheists around the Fallen world scramble to avoid the imminent universal desolation, it's gratifying to watch this international Empire of Antichrist being universally destroyed more of flesh and bone than of walls and stone. And there's absolutely nothing they can do to stop it. Their universal desolation is inevitable unto the culmination of Ar

2/23/2010 2:42:01 AM 3 Dave U. Random <anonym...@anonymitaet-im-inter.net>
There is a problem in MATLAB 2009a using divideint function.

Hello people. There is a problem when we use the divide param divideint to train with early stopping. Validation ratio is used in test set and test ratio is used in validation set. You can correct it modifying two lines in divideint.m file. Move line 107 to line 106 and vice-versa. Till next time! Rub=E9n.

2/19/2010 10:14:26 AM 0 =?ISO-8859-1?Q?Rub=E9n?= <ruben.saborido.infan...@gmail.com>
animation videos address

need some address for collecting animation images and videos for industrial automation

2/18/2010 10:52:07 AM 0 bajji <harishankar0...@gmail.com>
Call for Papers: The 2010 International Conference of Manufacturing Engineering and Engineering Management (ICMEEM 2010)

CFP: The 2010 International Conference of Manufacturing Engineering and Engineering Management (ICMEEM 2010) From: IAENG - International Association of Engineers http://www.iaeng.org/WCE2010/ICMEEM2010.html Draft Paper Submission Deadline: 6 March, 2010 Camera-Ready papers & Registration Deadline: 31 March, 2010 WCE 2010: London, U.K., 30 June -2 July, 2010 The conference ICMEEM=9210 is held under the World Congress on Engineering 2010. The WCE 2010 is organized by the International Association of Engineers (IAENG), a non-profit international association for the engineers and the

2/18/2010 9:42:43 AM 0 WCE 2008 <wce_2...@iaeng.org>
Mocazo-Integrated Marketing Solutions

Generate More Business through Mocazo Integrated Marketing Solution we do sms marketing solution e mail marketing and online activity=92s. We are mobile advertisers where depending on the customer requirement we devise campaigns for brand building, sales increase, footfalls etc

2/15/2010 11:07:03 AM 0 mani vanga <mani.vang...@gmail.com>
The Greater Good

-----BEGIN PGP SIGNED MESSAGE----- Have you noticed how the evildoers of the world are working toward the greater good, wholly regardless of their motives? This is because the power of God is unlimited; whereas, the power of the Fallen, sinful man is limited by the aforesaid. ___________________________________________________________ Friday December 21st 2012 AD is 1039 days away and counting. The final worldly pope BXVI is 30,256 days old and counting.

2/15/2010 9:02:08 AM 2 Anonymous <nob...@remailer.paranoici.org>
EasyNN-plus multicore

EasyNN-plus 13.0b runs at low priority on all the CPU's that it can find using up to 100% of each processor. That means complex networks converge much faster without slowing anything else down. Steve -- Neural Planner Software Ltd www.NPSL1.com EasyNN-plus. Neural Networks plus. www.easynn.com SwingNN. Forecast with Neural Networks. www.swingnn.com JustNN. Just Neural Networks. www.justnn.com

2/10/2010 4:37:10 PM 0 Stephen Wolstenholme <st...@tropheus.demon.co.uk>
CMOS Circuit Design openings with a leading firm in Bangalore, India

Hi, We have an excellent opening for our below requirement =96 Circuit Design with one of our leading and reputed clients into product development and research and development activities in Bangalore, India. Please find below the details: Custom Circuit Design -: Active hands on experience in CMOS Custom VLSI Circuit Design using industry leading tools (CADENCE preferred) Thorough understanding of CMOS technology and devices. Familiar with VHDL, schematics entry, circuit simulation (SPICE), custom layout, layout verification methodologies, design for yield and manufacturability,

2/10/2010 4:27:43 PM 0 Abdul Rasheed <abdul...@gmail.com>
TradeStation 8.8 with Owndata 2.8 free download

go to http://www.top-tradesoft.com/ TradeStation 8.8 with Owndata 2.8 free download Wave59 2.13 RT download free WaveFin - Wavelets for Finance download free WaveTrader - CycleTrader 2000 by Bryce Gilmore download free Wealth-Lab Developer 4.0 build 4 download free Wealth-Lab Developer 4.0.2 download free Wealth-Lab Developer 5.3 download free Wealth-Lab Pro 4.3.36 download free Wealth-Lab Pro 5.1 download free Window On WallStreet Day Trader 7.0 download free WinStock Pro v2.3.1 download free WinWawes32 1.3.25? download free WizeTrade 6.0 Pro for eSignal download free WizeTrade

2/10/2010 2:53:45 PM 7 "B. K." <xrpsgro...@gmail.com>
Data training with backpropagation

Hi, i would like to train network using backpropagation in matlab with large data where the network structure is 5 nodes for input 6 nodes hideen 4 nodes output. i should recognize data of four class ( A ,B,C,D ) some of our data are as follows: (A) (B) (C) (D) 1 1 1 1 1 1 1 1 -0.00281 -0.00276 -0.00248 0.019727 -0.000696 -0.002768 -0.001826 -0.001998 -0.09323 -0.09280 -0.09280 -0.090197 -0.093245 -0.093217 -0.093170 -0.0

2/9/2010 7:53:07 AM 3 phdtaiz <shawkialdub...@gmail.com>
eigenvector with the minimum eigenvalue, how to find a stable solution

Dear all, Given a real symmetric matrix A, the eigenvector, V_min, for the minimum eigenvalue, Lambda_min, need to be calculated. Surely there are various numerical algorithms to calculate it, but my impression is that if the minimum eigenvalue is very close to 0, the solution is not numerically stable (the resulting V_min might be very different from the ideal one). I am thinking if there is a way to rearrange the matrix A to be B, who has the same eigenvectors but A's V_min now becoming V_max, corresponding to its biggest eigenvalue, whose numerical solution is more

2/7/2010 4:32:20 PM 4 goodjeff <jiang1f...@gmail.com>
Call for Papers: The 2010 International Conference of Financial Engineering (ICFE 2010)

CFP: The 2010 International Conference of Financial Engineering (ICFE 2010) From: IAENG - International Association of Engineers http://www.iaeng.org/WCE2010/ICFE2010.html Draft Paper Submission Deadline: 6 March, 2010 Camera-Ready papers & Registration Deadline: 31 March, 2010 WCE 2010: London, U.K., 30 June -2 July, 2010 The conference ICFE=9210 is held under the World Congress on Engineering 2010. The WCE 2010 is organized by the International Association of Engineers (IAENG), a non-profit international association for the engineers and the computer scientists. Our congress co

2/5/2010 10:41:16 AM 0 IMECS 2008 <imecs_2...@iaeng.org>
How to choose, structure, format, massage inputs?

Is there a set of guidelines for optimally structuring inputs? I have a set of data but I could structure the inputs into the network in a multitude of ways. I could separate out boolean information, normalize numerical data... all sorts of stuff could be done. Is there a science to "prepping" the data for input? Thanks!

2/2/2010 5:45:00 PM 0 taotree <jshell...@gmail.com>
EasyNN-plus 13.0a

EasyNN-plus V13.0a is available from today. The main change is as follows:- Functions can be added to Nodes and Layers of nodes. They are executed when the nodes are processed while the network is learning the training information in the grid. Some functions can decrease the number of cycles needed for the network to converge. For example if a hidden layer function is added to multiply the activation it can improve learning speed by 2 or 3 times. -- Neural Planner Software Ltd www.NPSL1.com EasyNN-plus. Neural Networks plus. www.easynn.com SwingNN.

2/2/2010 3:39:31 PM 0 Stephen Wolstenholme <st...@tropheus.demon.co.uk>
Multiple Back-Propagation Version 2.1.3 is now available

Multiple Back-Propagation is a free software application (released under GPL v3 license) for training neural networks with the Back- Propagation and the Multiple Back-Propagation algorithms. Features: - Easy to use; - Highly configurable; - Fast training; - Provides RMS graphics during training; - Provides Output versus Desired output graphics during training for the training and test sets; - Generates C code for the trained networks; - Allows input sensitivity analysis; - Supports NVIDIA CUDA. You can get the latest version at: http://dit.ipg.pt/MBP or alternatively at:

2/2/2010 1:53:13 PM 0 Noel Lopes <noel.lopes....@gmail.com>
Scaled Conjugate Gradient in Netlab

Hi, I am using scg function of netlab to train a mlp network. I want to set the lambda and sigma parameter before training starts. The function is scg(f, x, options, gradf, varargin). Options is a vector of parameters. I am not sure which parameter is lambda and which is sigma. The help manual does not give information about this. How should I set the parameters for lambda and sigma? Many thanks David

2/2/2010 11:47:07 AM 0 David <dtian.ty...@googlemail.com>
Category (-1,1) Input: Standardize? Scale?

I see three choices: 1) Do nothing, input as -1,1 2) Scale to +=AD=AD - .58 (approx) the max curvature point of Sigmoid 3) Standardize to zero mean and unit variance The mantra is 3) but I'm not so sure I want a "bias" built into my Category. For example, if I have 100 cases where 99 are -1 and 1 is +1, I get mean =3D .09 var =3D .04 To get var =3D 1, I need to scale by 5x resulting in a mean of -4.9 Shifting to get zero mean, I get (-9.9, 0.1) for my category input where 99% of the time the input will be -9.9 and the remaining 1% will be 0.1 I know a lot of discussi

1/29/2010 9:02:54 PM 4 TomH488 <tom...@gmail.com>
pattern classification with mlp

hi everybody; i have a problem in training in mlp, i read posted about this but i dont know improve performance of may question, i have 6 feature and 10 class and i have 32 training sample , i use this code for training data but i have not a good result, just 11 sample classified corrected, input=r'; %32*6 output=T'; %32*10 net = newff(minmax(input),[10 10],{'losig' 'logsig'},'trainlm'); net.trainParam.epochs = 100; net.trainParam.goal = 1e-10; net = train(net,input,output); it train one epoch "TRAINLM, Epoch 1/100, MSE NaN/1e-010, Gradient NaN/1e-010 TRAINLM, Max

1/29/2010 6:54:52 PM 1 nasim <nasim.breeze112...@gmail.com>
Beyond Standardization: Representation

Scaling, Normalizing, Standardization are all techniques to, let me use the word, "balance," a network so that stable, accurate results are obtained. Mathematically, you are trying to reduce ill- conditioning resulting from interaction between very large numbers and small ones. I'm not trying for a precise mathematical definition here but just covering some intuition. Consider a problem with various inputs that classify into three groups, A, B, and C. Suppose they are Pressure, Temperature, and Density. For input case values, you have respectively, 100, 30, and 10 values for a

1/29/2010 6:45:33 PM 0 TomH488 <tom...@gmail.com>
Call for Papers: The 2010 International Conference of Applied and Engineering Mathematics (ICAEM 2010)

CFP: The 2010 International Conference of Applied and Engineering Mathematics (ICAEM 2010) From: IAENG - International Association of Engineers http://www.iaeng.org/WCE2010/ICAEM2010.html Draft Paper Submission Deadline: 6 March, 2010 Camera-Ready papers & Registration Deadline: 31 March, 2010 WCE 2010: London, U.K., 30 June -2 July, 2010 The conference ICAEM=9210 is held under the World Congress on Engineering 2010. The WCE 2010 is organized by the International Association of Engineers (IAENG), a non-profit international association for the engineers and the computer scientists

1/29/2010 10:21:16 AM 0 WCE 2008 <wce_2...@iaeng.org>
Can Someone please help me with my Computer stuff? 42570

I use the computer and the Net on a regular basis but I don't know a whole lot about it all, especially when it gets technical. I have an Acer Aspire AM5100-U5300A. I would like to speed this computer up, I would like a very good grasphics card without spending a fortune. Once I find out what I need I willo just go to ebay but I don't know what kind of meory, cards and such are compaitble with my machine. Help please. Virgina ghtpl@yahoo.com http://ipods-imacs.com E_/HyGFtrHAJiwDAKG=eMd,wz

1/27/2010 7:58:12 PM 0 Virgina Hernandez<gh...@yahoo.com>
High Level Encryption to Protect Your Patient and Research Data

High Level Encryption to Protect Your Patient and Research Data The Kyrptonite series of programs uses a high-level encryption algorithm. The algorithm and method was tested by the National Security Agency, and was one of the few programs originating outside of the agency considered for additional testing and evaluation. Due to the advanced algorithm of the program, this program is limited to sales inside the United States only due to evaluation by the Bureau of Industry and Security of the department of commerce. The Kryptonite-3 program allows you to have 3 keys that can be arra

1/26/2010 5:10:31 PM 0 Diana <d.kane...@yahoo.com>
Call for Book Chapters--Social Networks Analysis and Mining: foundations and applications by Springer Verlag

Call for Book Chapters--Social Networks Analysis and Mining: foundations and applications by Springer Verlag (Please accept our apologies if you receive multiple copies of this message) The study of social networks originated in social and business communities. In recent years, social network research has advanced significantly; the development of sophisticated techniques for Social Network Analysis and Mining (SNAM) has been highly influenced by the online social Web sites, email logs, phone logs and instant messaging systems, which are widely analyzed using graph theory

1/26/2010 5:22:53 AM 0 "tansel...@gmail.com" <tansel...@gmail.com>
SCface - Surveillance Cameras Face Database

Dear colleagues, We would like to inform about a new face database: SCface - Surveillance Cameras Face Database. The database is available free of charge to researchers, and obtainable through the procedure described at the SCface web-site: http://www.scface.org/ Here is a short description of the database, as presented on the above web-site: SCface is a database of static images of human faces. Images were taken in uncontrolled indoor environment using five video surveillance cameras of various qualities. Database contains 4160 static images (in visible and infrared spectru

1/25/2010 9:09:27 AM 0 Kresimir Delac <kde...@gmail.com>
Transfer functions -- sigmoid, Gaussian

Hello everyone. I am aware from some reading that the ability of neural networks to function as universal approximators does not depend on the non-linear transfer function used by each neuron. Rather it is a function of the network topology itself. That being said, I have come across some hints that transfer functions other than the standard sigmoid might have some utility. I seem to recall, looking through the documentation for the FANN package, that one could choose to use a Gaussian transfer function instead. I am guessing that changing transfer functions might allow differ

1/24/2010 11:15:22 PM 5 John Ladasky <lada...@my-deja.com>
The internal workings of the human neuron

What is the current scientific status of the idea that a human neuron could be a quantum computer? I would recall that this idea arose in the 1990's, but I have not been following the scientific discussion concerning that thought. kind regards, Mr Antti J. Ylikoski Helsinki, Finland, the E.U. (I posted this to the comp.ai newsgroup, which is moderated... The moderators rejected it. Have I become a persona non grata here?) Idem

1/22/2010 12:45:34 PM 4 Antti Ylikoski <...@vipunen.hut.fi>
BCPNNs

Is there a decent detailed description of Bayesian Confidence Propagation Neural Network (BCPNN) out there? I've been looking everywhere and I can find bits and pieces, but nothing that would provide me with enough information to actually implement one. Any help in finding a detailed description, or even better, existing source code, would be fantastic. Thanks

1/14/2010 5:19:11 PM 0 "Fredo" <fr...@hotmail.com>
Multiple Back-Propagation Version 2.1.2 is now available

Multiple Back-Propagation is a free software application (released under GPL v3 license) for training neural networks with the Back- Propagation and the Multiple Back-Propagation algorithms. The latest version of Multiple Back-Propagation can be downloaded at http://dit.ipg.pt/MBP/Download.aspx Version 2.1.2 is now available for download. This version supports CSV (Comma-separated values) files. CSV files can easily be generated by many programs such as Excel, thus CSV files are now the preferred method of input data (the other formats remain only for compatibility purposes) to Mult

1/7/2010 1:31:45 PM 1 Noel Lopes <noel.lopes....@gmail.com>
LOJ algorithm

Hi! Does anyone know details of that algorithm (the law of the jungle)? I found some information in http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.29.7532&rep=rep1&type=pdf but I need an explanation of every single step. One uses something called "winner history table". I could not find anything about it on the net, however. Maybe some of you may help me?

12/27/2009 1:40:56 PM 2 JackK <jackk...@gmail.com>
SOM, automatic cluster boundaries detection?

Hi, I've been searching a lot for methods which automatically detect clusters boundaries ( if any ) with SOM. The vast majority of references resemble on the analyst expertise to visually take the decision in the number of the clusters and where are their boundaries. One exception so far was this reference Determining Cluster Boundaries using Particle Swarm Optimization Anurag Sharma, and Christian W. Omlin Anyone has other references and comments about that?

12/22/2009 5:59:31 PM 5 joaocarlos <jcanist...@gmail.com>

Pages: 2









Newest Articles

=?ISO-8859-1?Q?File_output_and_the_degree_=28=B0=29_character?=
22 min. 38 sec. ago

Re: FindRoots?
23 min. 50 sec. ago

LinearModelFit
24 min. 1 sec. ago

Re: FindRoots?
24 min. 12 sec. ago

Printing problem of pdf created using movie15 Package
28 min. 40 sec. ago