If you find this page useful, please consider making a quick and secure donation (powered by PayPal) to keep it free!

Whiteboard

From The Thalesians

Jump to: navigation, search

2A9511BM0K0

http://en.wikipedia.org/wiki/Bias_of_an_estimator#Sample_variance

http://www.mathhelpforum.com/math-help/calculus/17731-solved-i-need-help-derive-d2y-dx2-integral-form.html

bjam --build-type=complete

http://www.nuovaelettronica.it/it/volume/index.cfm?ww_volume_id=19&ww_tipo=C

currency strategy by callum henderson

\def\verA{1}
\def\verB{2}
\def\ver{1}
\ifx\ver\verA
%do something for version 1
\fi
\ifx\ver\verB
%do something for version 2
\fi

http://blogs.msdn.com/joshpoley/archive/2008/01/24/custom-debugger-auto-expansion-tips.aspx

0         1         2         3         4         5         6         7
012345678901234567890123456789012345678901234567890123456789012345678901234

http://blogs.msdn.com/joshpoley/archive/2008/01/24/custom-debugger-auto-expansion-tips.aspx

http://www.manicai.net/comp/debugging/days/visualizer/

http://lifeofaprogrammergeek.blogspot.com/2008/05/cuda-development-in-ubuntu.html

http://www.google.com/apps/intl/en/business/outlook_sync.html

http://www.reimeika.ca/marco/prettyplots/

http://www.wlug.org.nz/MakefileHowto

http://groups.google.com/group/thalesians-gpus-in-finance-workshop

http://www.eggheadcafe.com/articles/20050511.asp — Visual Studio.NET: Cool Debugging Tricks

Matthew Proctor: http://notdennisbyrne.blogspot.com/2009/06/option-profit-and-loss-in-erlang.html

http://mags.acm.org/queue/20090203/?pg=12

http://www.vogella.de/articles/RichClientPlatform/article.html

http://home.datacomm.ch/paulsoderlind/Research/SafeHaven.pdf

http://qwafafew.org/

http://cognitrn.psych.indiana.edu/rgoldsto/complex/

http://hebb.mit.edu/courses/9.641/2002/lectures/

http://books.google.com/books?id=FltZ76gMmz8C&pg=PA238&lpg=PA238&dq=risk+appetite+ratios&source=bl&ots=rtko-fVBGM&sig=ZQiE53xBy2ISAzjcw4ozqH0Gfgk&hl=en&ei=eXq6Sb2RDJyD-AbbsdTHBA&sa=X&oi=book_result&resnum=10&ct=result#PPA247,M1

http://www.wilmott.com/messageview.cfm?catid=11&threadid=67610

http://www.fxoptionsguy.com/

http://www.freeos.com/guides/lsst/ch03sec02.html

http://www.cl.cam.ac.uk/~twh25/aai/

Hopfield model -- have there been any recent advances? Is it a good model of brain activity?

James Taylor Medical Devices Professional

In the process of doing my graduate work, applying various neural network approaches to clinical event detection, I mentioned the analogy between these networks and brain function, to a clinician. I didn't even suggest modeling brains. Needless to say I walked away crispy.

Bottom line: not only are these learning models absurdly simple and simplistic to be able to capture the behavior we know about, we know only a fraction of what goes on in a brain, blurry global averages, of what we need to know to model brain activity in a realistic way.

The engineers and modeling folks at best can build analogous structures that capture simplified behaviors in the neural structures. We see the lightning, but only know how to strike sparks.

Robert Solomon Independent Medical Practice Professional

It is difficult to mathematically model brain activity for a variety of reasons. One is the degree of complexity; there are billions of neurons and trillions of glial cells. The number of connections is literally astronomical. There are quantum effects in the brain. As diffusion tensor MRI maps of the brain improve, I think this will effect the static AI oriented mathematical models that are out there. The brain is qualitatively different than a massively parallel CPU,PC or even computer network.

Robert Solomon also suggests this expert on this topic: Ogan Gurel

Michael Will Director at Picodoc Corporation

While the Hopfield neural network is certainly interesting, it is really just a mathematical expression, and not intended as anything actually biological. Its neurons are binary, and it uses full connectivity (every neuron connected to EVERY other neuron). I would concur with what was said above about black box vs actual mechanism modeling.

Darin Takemoto Senior Research Scientist at Vertex Pharmaceuticals

The brain is an extremely complex organ with many functions assigned to specific regions. It is unlikely that a simple model such as Hopfield's could ever hope to model this, nor do I suspect it was ever intended to. I think the best one could hope to do would be to attempt to model a relatively simple subset of brain activity, such as say, processing of auditory or visual input, but even that is a long ways away as far as I can tell.

The furthest along we have come in modeling any sort of brain activity is probably attempting to model brain disorders. In these models the brain is considered a black box with inputs and outputs. The model is usually trained using in vivo or in vitro animal models of the disease in question. In this or any other computational modeling of disease, no attempt is made to understand in detail exactly what is going on, instead the focus is on the parameters that affect the disease state.

-=-

Hopfield model -- have there been any recent advances? Is it a good model of brain activity?

Clarification added 23 hours ago: Thank you very much for your answers so far!

I guess I should add a clarification.

My question is really in two parts.

(1) Has there been any recent work on the Hopfield model (and as I'm a mathematician / statistical physicist / computer scientist, I'm really looking at recent progress in these fields).

(2) Has there been any recent work specifically in trying to emulate some of the processes that take place in the human brain, particularly logical reasoning?

I certainly do appreciate that on it's own the model is too simplistic to represent the brain faithfully, I certainly agree with your

Neural Network Research - Is it still active? What are the hot topics?

Looks like most books/papers were published in the 80s / 90s. Am I wrong?

Are there any recent reviews or perhaps "Advances in..." kind of articles? Many thanks for all your help!

---

Kevin Hunt BA (Hons) MBCS CITP Information Technology and Services Professional

The Dorset Branch of the British Computer Society had a talk in February by Dr Christian Huyck on his neural agent leading towards AI. I've given the link (see about half way down the page) and you may want to contact the speaker to see if he can pass on any advice/comments. Links: http://www.dorset.bcs.org/recent.php

---

Rory Winston Senior Developer at HSBC

It still seems to be a pretty busy area - although as it might be classed as more mature these days, maybe more in terms of incremental improvements to the models than in any revolutionary breakthroughs. There seems to be a steady stream of NN-based papers coming through the ACM and IEEE research channels, for instance. Geoff Hinton gave a talk at Google a year or two ago on his current NN research, which you can find on YouTube. Links: http://www.amazon.com/Foreign-Exchange-Rate-Forecasting-Artificial-Internat...

Check out the "Advances" series by Springer: http://www.springer.com/computer/artificial/book/978-3-540-34628-9 (many of these papers may be freely available on arXiv)

Also, look at conference proceedings like SIGKDD or SIGEVO - there are a whole bunch of NN-related papers presented at conferences like that. You might also find some good papers in the IEEE and ACM journals on machine learning (I think the neural networks journal may have been folded into a more general machine learning one a while back). Finally, there are new books coming out , for instance I read this one last year:

http://www.amazon.com/Foreign-Exchange-Rate-Forecasting-Artificial-International-Operations/dp/0387717196

Unfortunately, I cant really recommend it, but Im sure there are a lot of better ones!

---

Dorina Grossu Group Initiator at Continuous Improvement Group

http://www.genomeweb.com/blog/combinations-permutations-and-sequencing Even more than before! Clarification added 19 hours ago: http://www.genomeweb.com/ucsd-researchers-use-proteomic-analysis-map-neural-proteins and another one!

http://www.mitpressjournals.org/doi/abs/10.1162/089892903321107891?journalCode=jocn

if [ 3 -ne 2 ] && [ 3 -ne 3 ] && [ 3 -ne 4 ]; then
    echo "NOT EQUAL"
fi

Learning Symbolic Rules Using Artificial Neural Networks

http://www.thalesians.com/finance/index.php/Events/Seminars/Template:Seminar

introduction to the theory of neural computation

http://www.springerlink.com/content/a032463l78255722/

http://www.scholarpedia.org/article/Hopfield_network

Sasha Cyganowski, Peter Kloeden, Jerzy Ombach. From Elementary Probability to Stochastic Differential Equations with MAPLE. Springer, 2002.

[edit]

How to add a new user when anonymous wiki access is not allowed?

http://www.amazon.co.uk/Beat-Forex-Dealer-Insiders-Exchange/dp/0470722088

http://cpmppi.blogspot.com/

http://www.bestexchange.net/

http://www.coderanch.com/t/429788/Java-General-advanced/java/Generified-singleton

http://www.forbes.com/2008/10/07/securities-quants-models-oped-cx_ss_1008shreve.html

http://www.moneyweek.com/investment-advice/how-to-invest/are-quant-funds-to-blame-for-volatility.aspx

http://www.charity-commission.gov.uk/publications/cc11.asp

http://wiki.kartbuilding.net/index.php/Trac_and_SVN

https://angel1.projectlocker.com/CodeTalents/CodeTalentsSite/trac/roadmap

http://foror.ru/articles/tapestry5/

http://forums.alfresco.com/en/viewtopic.php?t=856&highlight=hotspot

// here is some code...
for i := 0 to high(myArray) do 
begin
  writeln('Hi '+myArray[i]+'!');
end;
 
for i := high(myArray) downto 0 do 
begin
  writeln('Hi '+myArray[i]+'!');
end;

http://arxiv.org/abs/0705.0010

IDEA: Heatmaps

IDEA: Drawer

http://koichitamura.blogspot.com/2008/06/boostpython-examples.html

IDEA: The Psychology of Programming

IDEA: monitor dashboard / application

http://www.mathematik.com/Kolmogorov/

http://blogs.msdn.com/haibo_luo/archive/2007/10/16/5482940.aspx

http://gillesdaniel.com/papers/1997.Challet,Zhang.Emergence%20of%20cooperation%20and%20organization%20in%20an%20evolutionary%20game.pdf

http://www.biomedcentral.com/1471-2105/4/20/

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/325205 -- memoisation http://www.python.org/peps/pep-0318.html

http://www.cs.hmc.edu/qref/latex/qref#AMS -- AMS formatted proofs and theorems

http://gnosis.cx/publish/tech_index_cp.html -- Charming Python, a column by David Mertz

Calling super constructor in Python; multiple inheritance: http://mail.python.org/pipermail/python-list/2002-February/126059.html

CVS keyword substitution: http://ximbiot.com/cvs/manual/cvs-1.12.12/cvs_12.html

-- Date select : get only the date, time = midnight

declare @date1 smalldatetime

select @date1 = convert (smalldatetime, convert(varchar(10),getdate(),101) )

[edit]

Now in LaTeX!

I can now display $LaTex: \LaTeX$ on this page. To find out how, read this. But first, let me prove that I can do it and welcome my visitors with the famous Black-Scholes PDE:

$LaTex: \frac{\partial V}{\partial t} + \frac{1}{2}\sigma^2 S^2\frac{\partial^2 V}{\partial S^2} + rS\frac{\partial V}{\partial S} - rV = 0.$

http://answers.yahoo.com/question/index?qid=20071004080528AAYFDwt

http://www.codeproject.com/KB/recipes/MathieuMathParser.aspx?df=100&forumid=59609&exp=0&select=2173221

#pragma warning (disable: 4786)

http://blog.danbartels.com/archive/2004/10/06/210.aspx

http://support.microsoft.com/kb/178474

I too had a similar problem after installing Visual Studio.NET 2005 and converting a ASP.NET 1.1.4322 project to ASP.NET 2.0.50727 version Web Application Project.

First, I ensured that my .csproj file had the correct definitions like this:

This still did not work, so after casting about for a while,

I went to the Main Menu in VS.NET 2005 where I selected:

Tools > Options > Debugging > General and unchecked the "Enable Just My Code (Managed only)" Checkbox.

This final action enabled debugging in this project.

Hope this helps

Retrieved from "http://www.thalesians.com/finance/index.php/Whiteboard"