10 psychological experiments you must see

In this link: http://goo.gl/DFHJKq you can see some short description and videos on 10 most famous psychological experiments. It's interesting. worth to take a look. 

a possible way to analyse brain, my thesis!

-->

1- Hilary Putnam on left:  brain is an information processing system, the same as computers. He called it Computational theory of mind.

2- David Marr tri-Level Hypothesis: to understand an information processing system we should analyze it in three levels: Computationaly (what’s input and output, what's the problem), Algorithmic (how to solve it) and implementation(physical level).

Based on 1 and 2:  understanding the brain means, analysing all cognitive models in three levels of Marr.

For the first levelof Marr, Computational case, what we can do?

We can check if it is NP-Complete or not.

Let’s think its NP-Complete then it means it’s complety wrong? In other words our brain is not capable of  running such a model?

Nope, brain is much more smarter, it shrinks some parameters then solves the problem.

It means Instead of NP-Completeness we should check Prameterized complexity analysis.

Yes, That’s all. My thesis!

IBM SyNAPSE project

Awesome! They are going to build a new architecture for computers, completely far from today's von Neumann's ones.
This brain Inspired architecture will bring a new thinking, new softwares and new algorithms. Let's see what will happen. 

Glia! The mysterious parts of our brain

Glia, the cells that 85% of our brain are them.
What they do? Making the environment ready for neurons to do their job.
Just this? No. Recently, it is showed that the Astrocyte glia cells also infer processing information but not with electric signals as neurons do.
Wow! It's really amazing.
I don't know what will happen to neuroscience by new discovery, but at least in Artificial neural networks we can have both of these guys.
;) 

Artificial Psychology book

I've just finished taking a look at the book that my supervisor lend me: Artificial Psychology by Jay Friedenberg.
I can just say, it's Super Awesome, it's one of the most interesting books I've ever seen.
It discussed different topics in cognitive science such as learning, intelligence, consciousness, creativity, emotions and many other things then talked about possible ways of modeling and also previous and ongoing work in these fields. It just give the general ideas, so it's much easier to read and more fun.

Thanks Jay.
Thanks Todd.

Mark Changizi TED talk about illusions

Interesting! His idea: Our brain predict everything a small proportion of a second in advance, That's why we see illusions.

Happy Nowruz, Iranian New Year

HAPPY NEW YEAR ;)

Blue Brain Project and the 1.3 billion dollar

Now, it's  a while, that my love, the Blue Brain Project has got 1.3 billion dollars from the European Commission to finish it's work:

A very detailed simulation of human brain, a model that is hoped to work near to our brains. A new intelligent creature!!!

Markram, the head of the project, now seems has everything, money, supercomputers, good team, attention from the media and in short, what he needs. But for sure it needs extreme efforts to  get finished in the proposed  year, 10 years later.

I'm impationtly waiting to see their success whitin 10 years. Luck with you guys.

For more information about this project just check their official page that now  almost after being billioner changed to Human Brain Porject. ;)

Parametrized Complexity and Cognitive Science and Everything

Parametrized Computational Complexity, Who is this guy? a good guy or a bad guy,Let's see.

In Classic Computational Complexity, we say this algorithm is in NP, so kick it out. It's rubbish.

But Parametrized Computational Complexity says:

keep calm classic! The algorithm maybe is in NP but it's nature is p some damn parameters made it NP. So we can control these bad parameters(source of intractability), keeping them small and have a nice p algorithm.

The cognitive models I talked about in last post and I said they are considered rubbish, they may be acceptable by parametrized Complexity. The FPT(Fixed parameter Tractable) which hopefully I'll talk about in next posts is a tool for checking tractability in the view of parametrized Computational Complexity.

Hope to see you soon guys ;)

Computational Intractibility, The wet blanket!

Up to know, psychologists has introduced so many cognitive models, models that explain how our mind works.
BUT
What if our mind is just a Turing Machine, then are these models will work on our brain or maybe they are just some imaginations, and our brain never will be able to run them.

Oh, no, in this case all these models are just a rubbish?  Poor the ones who put a lot of effort on a model without checking its computational tractibiliy first.

Even if our brain is not a Turing machine and be superiour to it, still we need to check the tractibility of all models, because AI may need and can implement these models some day. And at least we know AI is on Turing Machines.

SO LET'S GO TOWARD CHECKING COMPUTATTIONAL TRACTIBILITY OF COGNITIVE MODELS, AND REJECT SOME FAMOUS COGNITIVE MODELS!!!

I'll post more about this issue, Have fun  ;)

Could our computers memories like our brains?

yesterday, I was thinking about swimming.  It lead me to think about my friend, Parvaneh(it means butterfly in Persian) who talked about swimming classes. suddenly, I thought she is professional in butterfly swimming.
After a while, I understood I made a mistake. Because of her name, I thought she knew butterfly swimming :D

I think we could have computers with a memory like ours. A fast efficient memory which may make mistakes.
It seems, this kind of model for saving information matches Memristor based hardwares(I donno much about Memristor).

 I searched the net to find if there is any similar computers, but no success yet.
I tried to write the relation between words which came to my mind (butterfly, swimming, my fried talk) in a flowchart like graph. but I don't know how true it is.

I always think by myself: as there should be a theory of everything in physics; it is not weird to have such a thing in neuroscience; hmmm, a set of rules or even one rule that can explain everything happens in our brains. It'll be fantastic.

Reconstructing brain vision

I have just read about the amazing work done by Berkeley neuroscienists: capturing visual activity in brain. the above video shows the result. It is wonderful. who can believe one day scientists could see what you see through your brain. more information is available here.

Earth Day

some tips about Batteries

Batteries that are thrown away produce most of the heavy metals. They can be harmful to humans and wildlife.

• Turn off the toys and games that use batteries when you are not playing with them. That makes the batteries last longer, and you won't need as many of them.

• Buy rechargeable batteries and a recharger.

• Check to see if your device has a "Battery Save" option. If so, you can turn this on to increase the life of your battery

• We should all take batteries to a toxic waste disposal area if it is possible.

From: WikiHow & EnergyQuest

Iranian New year starts on Nowruz

Iranian new year, Nowruz will start a few hours later with a lot of fantastic traditions as seven S'es table, I hope this year be a nice one for all people around the world^_^

I hope every day of your life be Nowruz and you be happy and successful on Nowruzs :)

Technically, the glass is always full :)

Dual ... A keyword for modeling the brain

Dual representation
Dual working
Dual neurons
Dual ....
These days it seems Brain modelers are more interested in using dual things in their models and interestingly by using this word, they have got good results.
Here are some examples of some works around the world:

• USA: Professor Ron Sun's cognitive architecture, He's used dual representation of the brain to build a cognitive modeling software, CLARION. He was so successful in creating creativity in hardware(one of my two last posts was about this.)
  

• Turkey: Professor A. Khashman's published an article recently (A Khashman - Neural Networks, Dec 2010 - Elsevier). He's used a novel emotional dual neural network architecture in face recognition and he got much better results than other models.
  

• Iran: Professor Majid Nili AhmadAbadi: Recently, In AmirKabir cognitive science seminars he talked about good results comes by combining two models and used a dual... I searched the article he talked about but I didn't find it. This article seems to be another successful dual model he works on.
  

I've searched the net to find a theory about what exactly this dual is, and I found Dual-Coding theory.
Maybe the parallel nature of processes in the brain and the relation between different parts of the brain makes this dual works well.
Maybe triplex, foursome, quintuplet, .... works better, betterer,bettererer,... :)

keep evolving

We've all heard of Moore's Law, which states that the number of transistors placed inexpensively on an integrated circuit doubles approximately every two years. Well, since this issue covers human computation, let's have a look at our own biological hardware. How are we doing compared to a previous model? Keep evolving, folks...

—James Stanier

Human Creativity in hardwares

You can't imagine how much I was excited when I saw this in ACM tech news. The title of this, is:

New Psychology Theory Enables Computers To Mimic Human Creativity at Rensselaer Polytechnic Institute



Could you imagine a computer which is able to answer questions need creativity?
like this question:

A dealer in antique coins gets an offer to buy a beautiful bronze coin. The coin has an emperor’s head on one side and the date “544 B.C.” stamped on the other. The dealer examines the coin, but instead of buying it, he calls the police. Why?

Ron Sun and Sèbastien Hèlie in their recent paper (appeared in the July edition of Psychological Review) introduced “Explicit-Implicit Interaction Theory,”
Sun has developed CLARION computer program(like a cognitive system) based on this theory.
Comparing the performance of the CLARION model with results from previous human trials – including tests involving the coin question was amazing:

Clarion, answered these kinds of questions with a great accuracy correctly in several aspects of problem solving.

The researchers say the theory, which explains how humans solve problems creatively, could provide a blueprint to building AI systems that perform tasks like humans.

If you remember, one of my recent posts was about moral judgment: "killing one to save five". ClARION is a very useful software for simulating these kind of tasks, I can't wait to see the results. I am eager to know how much CLARION's moral judgment is like ours.

You can see the RPI news about this, here.

:) wireless technology and the future of birds

it's a bit freaky with this wireless technology!