FRIHOST • FORUMS • FAQ • TOS • BLOGS • DIRECTORY
You are invited to Log in or Register a Frihost Account!

Sudoku: a challenge for your brain; beyond computers?

   Frihost Forum Index -> Science -> General Science
 


SonLight
The game of Sudoku (9x9 grid, fill in with 1 to 9) has been studied recently, and it was found that when humans solve it, they use quite sophisticated forms of associative memory. That seems to account for the observed fact that the game is a good one for exercising the brain.

It's not really beyond computers, of course, but today computers usually use a much different method than humans to solve the game. and it is believed that if they can be programmed to solve the game like a human, it could lead to a big advance in artificial intelligence.

The article is at:
http://www.abc.net.au/science/news/stories/2006/1737840.htm

For a good game of sudoku, with helpful solving assistance, you may want to check out brainbashers ( http://www.brainbashers.com/sudoku.asp ).
EanofAthenasPrime
Well, the way a human brain works is through extrapolating physical memory and cellular micro capacitation to do complex calculations. This essentially means that brains do math by recalling a memory of what steps to do the math, then do simple recalled calculation algorithms step by step to accomplish the calculation. This method is very slow so computers have the advantage. Computer's methods are millions of bits (capicatators with a voltage equal 1, non-charged capacitators=0) using standard algorithms to accomplish calculations much faster.
nilsmo
SonLight wrote:
[humans] use quite sophisticated forms of associative memory. ... it is believed that if they can be programmed to solve the game like a human


Associative memory? Sounds like learning (associating memories with other memories). So the article is a lot of fluff. But they mentioned that there has been some work in Hopfield's model - anyone know what that is?
Nyizsa
We, humans, really think about these riddles, while computers try every possibility. You could also solve it this way, say, in 3 years. Computers do it in a matter of seconds.
Computers are not intelligent, they are fast.
Moonspider
EanofAthenasPrime wrote:
Well, the way a human brain works is through extrapolating physical memory and cellular micro capacitation to do complex calculations. This essentially means that brains do math by recalling a memory of what steps to do the math, then do simple recalled calculation algorithms step by step to accomplish the calculation. This method is very slow so computers have the advantage. Computer's methods are millions of bits (capicatators with a voltage equal 1, non-charged capacitators=0) using standard algorithms to accomplish calculations much faster.


True, but imagaine the complex calculations that go into simple tasks, such as driving a car or simply catching a baseball. The human brain does all these transparently and with remarkable speed.

Respectfully,
M
polis
EanofAthenasPrime wrote:
Well, the way a human brain works is through extrapolating physical memory and cellular micro capacitation to do complex calculations. This essentially means that brains do math by recalling a memory of what steps to do the math, then do simple recalled calculation algorithms step by step to accomplish the calculation. This method is very slow so computers have the advantage. Computer's methods are millions of bits (capicatators with a voltage equal 1, non-charged capacitators=0) using standard algorithms to accomplish calculations much faster.


just one thing:

capicatators with a voltage over 5 volts equal 1, below 5 volts capacitators=0

Back to the subject:

What about instinct? How does sea turtles know that they must go to the sea as soon as they are born instead staying at the beach?

Is that intelligence?
EanofAthenasPrime
No, that is evolution. All of the sea turtles who did not have this compulsion to go towards the sea died, and those that did remain today.
polis
EanofAthenasPrime wrote:
No, that is evolution. All of the sea turtles who did not have this compulsion to go towards the sea died, and those that did remain today.


I think you didn't get my question.

How do they know that they have to go to the sea?
EanofAthenasPrime
polis wrote:
EanofAthenasPrime wrote:
No, that is evolution. All of the sea turtles who did not have this compulsion to go towards the sea died, and those that did remain today.


I think you didn't get my question.

How do they know that they have to go to the sea?


Evolution. I already told you. Some brains did not like the ocean, some did. The ones that did survived, the rest died. The living ones lived to pass the "love of the ocean" to their offspring. The one's that did not like the ocean did not live to have offspring. Its that simple.

This happened a long time ago, so now all sea turtles love the ocean.
polis
Honestly I think it has nothing to do with evolution.

Evolution explain why the instinct is there, but doesn't explain how it works. That's the whole point.
EanofAthenasPrime
polis wrote:
Honestly I think it has nothing to do with evolution.

Evolution explain why the instinct is there, but doesn't explain how it works. That's the whole point.


The theory of evolution does not, but science does. And it has much to do with evolution.
qsmith
It can't be all that difficult - Almost tempted to write a program to solve it quickly

Also wanted to do chess but that one is a bit Shocked bigger.

anyone want to join in?
Wynand
Moonspider wrote:


True, but imagaine the complex calculations that go into simple tasks, such as driving a car or simply catching a baseball. The human brain does all these transparently and with remarkable speed.



The brain does not need to do complex calculations to be driving a car or catching a baseball because this is done through a process of estimation, learning to estimate correctly and trail and error.

With calculations I mean number crunching.

When you for the first time try to catch a ball, you estimate how fast the ball is coming to you. You then learn through trial and error how fast you need to move your hand to catch the ball. If you do not catch the ball because you were to slow you will move faster next time. In this process the brain did not do any subtraction or multiplication.
EanofAthenasPrime
Wynand wrote:
Moonspider wrote:


True, but imagaine the complex calculations that go into simple tasks, such as driving a car or simply catching a baseball. The human brain does all these transparently and with remarkable speed.



The brain does not need to do complex calculations to be driving a car or catching a baseball because this is done through a process of estimation, learning to estimate correctly and trail and error.

With calculations I mean number crunching.

When you for the first time try to catch a ball, you estimate how fast the ball is coming to you. You then learn through trial and error how fast you need to move your hand to catch the ball. If you do not catch the ball because you were to slow you will move faster next time. In this process the brain did not do any subtraction or multiplication.

I'm sure it does plenty of subtraction and multiplication to access the memory strands. However, if what you are meaning is that is uses computer like algorithms and trigonometric formulae, then you are right.
ptolomeo
Everything OK with the mental process, but it is nothing new. As sudoku is very new and poplular now we have a man writing a paper about the mental process involved in solving it and recurring to the seller thing of artificial intelligence. Someone mentioned about driving a car, there is a lot about images brain recognition there that computers still can not do, at least not as human do. The interesting thing in Sudoku is its mathematics:

-How many numbers do you have to put initially to assure that there is a unique solution?
- Is there always a solution given a compatible set of initial numbers?
- Does the minimal number of numbers depend on the place they are placed initially?
- What are the minimal number of steps to solve the game?

I think these are the hard question (the artificial intelligence question is also hard but not limited to sudoku, the thing that makes sukdoku unique is this type of mathematical difficulties).
Does someone knows something about the math of Sudoku?
marrs
Its a Chinese game puzzle I think and we alll know they are very simplistic people with a very clean and effective way of living and this just shows us a bit of a different way of thinking
gong_05sk
i can finish it in 15 minute. it very wonderful...i'm relax....
EanofAthenasPrime
I don't get how to play. Evil or Very Mad
SonLight
EanofAthenasPrime wrote:
I don't get how to play. Evil or Very Mad


Look it up at Brain Bashers ( http://www.brainbashers.com/sudoku.asp ). They have some good explanations and tools to simplify play. It is basically a logic puzzle, where each position can only contain a specific digit. Each box of 9, each row, and each column must contain all nine digits with none repeated. If you find the number 1 already in row 1 and row 3, for example, then the 1 in row 2 must be in the other box in that row. Try a few easy ones that can be worked just by scanning rows and columns, then see if you can find logic rules to handle the harder cases, or read the help pages.

Once you get to harder levels, you need to mark all possible digits in each remaining square. I find this too tedious to do manually, but with BrainBasher's sudoku assistant, the game is challenging and reasonably paced.
EanofAthenasPrime
Thanx! this is the stuff my brain has to handle right now tho (overload):

Code:
dw=display_get_width()
dh=display_get_height()
if shape=0 {
if mouse_x>dw-74 and mouse_y>36 and mouse_x<dw-19 and mouse_y<55 {newxx=get_integer("X:",xx) xx=newxx
if xx=0 {xx+=1 instance_create(x,y,NoZero)} global.floors[geti,0]=xx}
if mouse_x>dw-74 and mouse_y>73 and mouse_x<dw-19 and mouse_y<92 {newyy=get_integer("Y:",yy) yy=newyy
if yy=0 {yy+=1 instance_create(x,y,NoZero)} global.floors[geti,1]=yy}
if mouse_x>dw-74 and mouse_y>108 and mouse_x<dw-19 and mouse_y<129 {newzz=get_integer("Z:",zz) zz=newzz
global.floors[geti,2]=zz}
if mouse_x>dw-74 and mouse_y>144 and mouse_x<dw-19 and mouse_y<163 { getex=get_integer("Model index-",tempgetex)
tempgetex=getex
if (geti mod 2)=0 global.floors[geti,2]=round(global.floors[geti,2])+getex/1000
if (geti mod 2)=1 global.floors[geti-1,2]=round(global.floors[geti-1,2])+getex/1000
}
if mouse_x>dw-74 and mouse_y>181 and mouse_x<dw-19 and mouse_y<200 {getuvX=get_string("Horizontal texture coordinate:",tempgetuvX)
uvX=real(getuvX) tempgetuvX=uvX
if (geti mod 2)=0 global.floors[geti,0]=round(global.floors[geti,0])+real(uvX)/1000
if (geti mod 2)=1 global.floors[geti-1,0]=round(global.floors[geti-1,0])+real(uvX)/1000
}
if mouse_x>dw-74 and mouse_y>216 and mouse_x<dw-19 and mouse_y<236 { getuvY=get_string("Vertical texture coordinate:",tempgetuvY)
uvY=real(getuvY) tempgetuvY=uvY
if (geti mod 2)=0 global.floors[geti,1]=round(global.floors[geti,1])+real(uvY)/1000
if (geti mod 2)=1 global.floors[geti-1,1]=round(global.floors[geti-1,1])+real(uvY)/1000}}
if shape=1 {
if mouse_x>dw-74 and mouse_y>36 and mouse_x<dw-19 and mouse_y<55 {newxx=get_integer("X:",xx) xx=newxx
if xx=0 {xx+=1 instance_create(x,y,NoZero)} global.wall[geti,0]=xx}
if mouse_x>dw-74 and mouse_y>73 and mouse_x<dw-19 and mouse_y<92 {newyy=get_integer("Y:",yy) yy=newyy
if yy=0 {yy+=1 instance_create(x,y,NoZero)} global.wall[geti,1]=yy}
if mouse_x>dw-74 and mouse_y>108 and mouse_x<dw-19 and mouse_y<129 {newzz=get_integer("Z:",zz) zz=newzz
global.wall[geti,2]=zz}
if mouse_x>dw-74 and mouse_y>144 and mouse_x<dw-19 and mouse_y<163 {tempgetex=getex getex=get_integer("Model index-",tempgetex)

if (geti mod 2)=0 global.wall[geti,2]=round(global.wall[geti,2])+getex/1000
if (geti mod 2)=1 global.wall[geti-1,2]=round(global.wall[geti-1,2])+getex/1000
}
if mouse_x>dw-74 and mouse_y>181 and mouse_x<dw-19 and mouse_y<200 {getuvX=get_string("Horizontal texture coordinate:",(global.wall[geti,0]-round(global.wall[geti,0])))
uvX=real(getuvX)
global.wall[geti,0]=round(global.wall[geti,0])+real(uvX)/1000}
if mouse_x>dw-74 and mouse_y>216 and mouse_x<dw-19 and mouse_y<236 {getuvY=get_string("Vertical texture coordinate:",(global.wall[geti,1]-round(global.wall[geti,1])))
uvY=real(getuvY)
global.wall[geti,1]=round(global.wall[geti,1])+real(uvY)/1000}}
if shape=2 {
if mouse_x>dw-74 and mouse_y>36 and mouse_x<dw-19 and mouse_y<55 {newxx=get_integer("X:",xx) xx=newxx
addxx=(global.triangle[geti,0]-round(global.triangle[geti,0]))*1000
if xx=0 {xx+=1 instance_create(x,y,NoZero)} global.triangle[geti,0]=round(xx)+addxx}
if mouse_x>dw-74 and mouse_y>73 and mouse_x<dw-19 and mouse_y<92 {newyy=get_integer("Y:",yy) yy=newyy
addyy=(global.triangle[geti,1]-round(global.triangle[geti,1]))*1000
if yy=0 {yy+=1 instance_create(x,y,NoZero)} global.triangle[geti,1]=round(yy)+addyy}
if mouse_x>dw-74 and mouse_y>108 and mouse_x<dw-19 and mouse_y<129 {newzz=get_integer("Z:",zz) zz=newzz
addzz=(global.triangle[geti,2]-round(global.triangle[geti,2]))
global.triangle[geti,2]=round(zz)+addzz}
if mouse_x>dw-74 and mouse_y>144 and mouse_x<dw-19 and mouse_y<163 {tempgetex=getex getex=get_integer("Model index-",tempgetex)

if (geti mod 3)=0 global.triangle[geti,2]=round(global.triangle[geti,2])+getex/1000
if (geti mod 3)=1 global.triangle[geti-1,2]=round(global.triangle[geti-1,2])+getex/1000
if (geti mod 3)=2 global.triangle[geti-2,2]=round(global.triangle[geti-2,2])+getex/1000}
if mouse_x>dw-74 and mouse_y>181 and mouse_x<dw-19 and mouse_y<200 {getuvX=get_string("Horizontal texture coordinate:",(global.triangle[geti,0]-round(global.triangle[geti,0])))
uvX=real(getuvX)
global.triangle[geti,0]=round(global.triangle[geti,0])+real(uvX)/1000}
if mouse_x>dw-74 and mouse_y>216 and mouse_x<dw-19 and mouse_y<236 {getuvY=get_string("Vertical texture coordinate:",(global.triangle[geti,1]-round(global.triangle[geti,1])))
uvY=real(getuvY)
global.triangle[geti,1]=round(global.triangle[geti,1])+real(uvY)/1000}}


That is just the code that modifies the properties of a point in a 3d modeling program I am making...
benjmd
I think it should be clarified that Sudoku is not math. The puzzles are indeed fun to do and give your brain a little exercise, but there is no math involved. There are no mathematical operations necessary to complete the puzzle at all. If you can count from 1 to 9, then you have accomplished the "math" part of the puzzle. The rest of it is a basic form of logic puzzle.
ptolomeo
I partially disagree with benjmd. While it is true that you should not know anything of mathematica to play sudoku (the only requisite to play it is being able to recognize 9 different symbols), I think it is full of wonderful mathematics. As I posted before:

Quote:

The interesting thing in Sudoku is its mathematics:

-How many numbers do you have to put initially to assure that there is a unique solution?
- Is there always a solution given a compatible set of initial numbers?
- Does the minimal number of numbers depend on the place they are placed initially?
- What are the minimal number of steps to solve the game?


These and related (still open) questions make sudoku interesting.
paul_indo
Sodoku is fun but it gets a bit boring for me after a while.

It is such a repetitive process to solve the puzzles. I prefer a game like chess.
Reply to topic    Frihost Forum Index -> Science -> General Science

FRIHOST HOME | FAQ | TOS | ABOUT US | CONTACT US | SITE MAP
© 2005-2007 Frihost, forums powered by phpBB.