Advertisement

Potentual new concept

Started by September 10, 2015 03:07 PM
98 comments, last by TheComet 8 years, 11 months ago
Hi all,
I would like to introduce an abstract Concept into the A.I. and Gaming community.
My only qualification is perception.I precieved the idea of a new concept there fore am qualified.
This being said;,I, really need a professional to understand,interpite,develope,and verify it's
potentual.
My belief is that it will introduce logic,and relitivity to programming;,Not,As a fast action,but,as an efficient over all learning process.
Just as A.I. learns with repatition this could be used to relate logics,and relations over time.
My introduction to the abstract concept begins with the observations which lead me to it.
This is important to show the facts from the abstract usses.
1. Parrallel resistance
A. effectiveParll_restance= 1 / ((1/r1)+(1/r2)+(1/r3)---)
a1. effectiveParll_restance is always less than least resistor.
a2. Before the last {1/} the sum is always intuativly a unique extension to the vector length. ie. .5+.25+.25 +.125=1.125
a3 The {{ effective_resistance is relitive to the lost_resistance and the total_resistance ie; effective_resistance = Total +(lost*-1) }} ie; 0.625= 1.125+((.25+.25)*-1) == .5 +.125
a4. Abstract:
if resistance is an asc(code) ie:asc(*) is 49, then the total of "country United States" is relitive to " United " = { "country United States" + (("country +States" )*-1)

B. Parrallel resistance program_loop format.
b1.
Stream$="800 96 50 12"
total=val(Stream$,x)
loop
x=x+1
r=val(Stream$,x)
total=(1/ (r*(1/total)+1)))*r
next loop
note total=8.69565217 == 1/((1/800)+(1/96)+(1/50)+(1/12))
ABSTRACT:
b1a. Observe the percentage relation r*%= effective_total
b1b .Observe A= atn(%)
A1
A2
A3
final A =A3 Note that r(n)= total(n)*(1/tan(A(n))) or T(n)=R(n)*(1/tan(A(n))
final total=final total
This is where logic_relative_addressing and relative_logic comes into play.
Abstract::
A Abstract relitive logic
1. P>0 .5
2. P<0 * .25
3. H=0 .136
4. H>0 * .33P<0
5. P=1 .4
total = 1.616
6. Abstract relitive logic mixer : ( {1.616 - 1.036 = .58} *-1) + relitive total ==
13 == " P<0 , H>0 " == "quadrent 2"

B Abstract relitive addressing
1 Page unique_code = 0.02425
2 Options unique_code total = 0.325
3 address is 0.30075= 0.325+(0.02425*-1)
4 action== goto "[";"0.30075";"]" == goto [0.30075]
'***************************
[0.30075]
open "page25.txt"
return

C Note the action is based on relations of content not on the content. ie; the action is assigned thru the relations that define it.

Heavy concept?? yes!
No one said dealing with logic,relitivity,reasoning,and selfawarness was simple.BUT YOU NEED TO START SOME WHERE. To understand it.

I believe this to be the starting place.
I need help proving it.
This is only an incomplete glimps at the detail.
'********************************************************************

SHOWING WHERE T AND A ENTER THE EQUATION::

Pr=1e100
{T(1),A(1)} {T(2),A(2)} {T(3),A(3)} {T(4),A(4)} Not in Formula form
Pr=(1/1e100) (1/800) + (1/96) + (1/50) + (1/12)
'********************************************************************

Pr=1e100 'initialize 'Abstract T
T=Pr

A=45 'initialize { relativity }

{ Pr) {T}
%= 1 / ((800*(1/1e100)) +1)
Angle=atn(%) 'Abstract A(0) -- Angle

T=800*tan(A) 'Abstract T(0) Pr(1)=T -- total(0)

T(1)=800
A(1)= 1 / ((800*(1/1e100)) +1)
Angle=atn(%) 'Abstract A(0) -- Angle

T(0)=800 'Abstract T(0) == Pr(0)=T -- total (0)
A(0)=45

{{ 'initialize complete}}'

'@@@@@@@@@@@@@@@@@@@
[loop]{1}
{ Pr} {T}
'''''/%= 1 / ((800*(1/800)) +1)
''''/ Angle=atn(%) 'Abstract A(0) -- Angle

T=800*tan(A) 'Abstract T(0) Pr(0)=T -- total T in Program form
T(1)=800
%(1)= 1 / ((800*(1/800)) +1)
Angle=atn(%) 'Abstract A(1) -- Angle A in Program form

T(1)=400
A(1)=26.565051177077989351572193720453

next [loop]
'@@@@@@@@@@@@@@@@@@@
[loop]{2}
{ Pr} {T}
''''/%= 1 / ((96*(1/800)) +1)
''''/Angle=atn(%) 'Abstract A(2) -- Angle

T=96*tan(A) 'Abstract T(2) Pr(2)=T -- total
T(2)=85.714285714285714285714285714286
%(1)= 1 / ((96*(1/800)) +1)
Angle=atn(%) 'Abstract A(2) -- Angle

T(2)=400
A(2)=41.760299703897870309452138780517

next [loop]
'***********************************************************************************************
With T and A in the equation it is concievable to back_track the process. ie; T=Pr*tan(A) and Pr=T*(1/tan(A))

'@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
Now the hard part..
This is the relitivity system.The SYSTEM that makes the Final_Code {FCode} relavent.

Pi=atn(1)*4:Mdeg=(180/Pi):Mrdn=(Pi/180)
DL$="800 96 50 12 3 7 18 21 41 45 96 50 12 3 7 18 800 96 50 12 3 7 18 21 41 21 41 45 96 50 12 3 7"
Pr=val(word$(DL$,1)):F=Pr
Pr=1e-100
for j=1 to Lng(DL$)
R=val(word$(DL$,j))
FCode=FCode+((1/R)+(1/A))
gosub
gosub [Decode]
next j
print eDL$
print FCode
end



'print R
c=abs(1/(R*(Pi^2)))
c1=c
Pr=F
A=atn((1/((R*(1/F))+1)))*Mdeg
L=(A*(1/(Pi^2)))
F1=1/(((L*c)^.5)*(2*Pi))
Pr=R*tan(A*Mrdn)
'print Pr
return

[Decode]
print R
c=abs(1/(R*(Pi^2)))
c2=c
F2=1/(((L*c)^.5)*(2*Pi))
'L2=((1/F2)/(2*Pi))/c2
F=(1/((1/tan(A*Mrdn))-1))*R
A2=L*(Pi^2)
D=Pr*(1/tan(A*Mrdn)):eDL$=eDL$;str$(int(D));" "
'print tab(16);D;" ";F1+(F*-1);" ";A2+(A*-1)
return
'@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@

The relivence of the FCode is the logic of the system.
Rule is {relitive to the system} FCode of "one apple = 35 cars" = only "one apple = 35 cars"
Therefore box_Fcode is the same as box_"one apple = 35 cars" ;, But , box_Fcode + box_Fcode2 is _"one apple = 35 cars" + box_something else. ie; "P<0" + "H<0"= "Quadrent 2"
How???? assigned logic --- if Code(1) + Code(2)= 0.73433 then GOTO= { Code= (Code(1) + Code(2) +Code(3) ) + ((Code(1) + Code(2))*-1) }
if GOTO= ? then print "Quadrent 2"

Now where we used resistance before;,We, can use coded logics and relate them as effective_Logic= 1/ ( (1/EL1) + (1/EL2) + (1/EL3) --)) or Gain=(Total+(Loss*-1))

The difference is that these are assigned abstract relitive relations {relitive to the rule of the abstract System}

Now we can assign logics,relations,relitivity,reasoning,creativity, ect -- as temporary or constants.
At {LEAST} it is a start at understanding how it works.

If interested My e-mail is [deleted by moderator]
Garbage in, garbage out.
Advertisement

You used alot of jargon, but didn't really say anything. Saying something in a confusing way doesn't make it a smarter idea.

You might have something really valuable to say, but until you can communicate it clearly, you probably haven't refined it enough in your own head.

In English, what does this new concept do? How does it do it? I'm not interested yet in how it's implemented, just the general plain-English reason for why I should care. That's where you have to start.

Can you explain, in one jargon-free paragraph, what it does and how it (conceptually/abstractly) does it? The general idea, I mean.

Sorry guys.

My communication is poor.

Let me work on it and get back to you.

thanks for the input.

I hope this helps.

For lack of a better name; let's call this - Logic Relative Addressing.
The unique code of logic is the logic address;, As pressing key B on a key pad takes you to memory biniary address B start or an icon opens a file.
Let say these are known truths.
They are predefined and techniclly manually set in memory.
But what if we wanted to get a known relitive truth based on more than one key or icon and past knowldge ? {relalitive influence}
Now the storage and recall has to be relitive,automatic,accurate,munipulitive,and versital.
So this is about using past relations with present conditions or relations to get a new potentual relation and if good storing it as an experianced known truth.
Thus giving relitive learning.
Primarialy it's about working with a form of relitivity and exsperenced relitive storage of logic.
In a game maybe the next punch is a kill punch.
past logic tells you this.
present logic tells you this is a friend that is not in control of his actions.
Do you kill?
Do you controll?
How do you change your actions to accomplish this?
Have you done this before?
Are there a simular situations {logics} you can combined to answer this?
What are the new alterations to your actions.
Now act!!
Relitivety gives you near instant access to the conclution.

This, unfortunately, does not clarify anything.

Advertisement

Ok.

I did say I wasn't qualified.

I don't want to spam so I'll leave it as is.

if you decided it's worth anything.I gave my email.

Thanks for your comment and time.

Donnie.

Markov chains?

Ok.

I did say I wasn't qualified.

I don't want to spam so I'll leave it as is.

if you decided it's worth anything.I gave my email.

Thanks for your comment and time.

Donnie.

It has Nothing to do with being qualified, zero of your content made ANY sense whatsoever, i don't think anyone here even understands what you're trying to talk about, if you can't communicate your idea clearly then there's no idea truly formalized in your head, remove all the technical crap and explain what it is about simply in 2 to 5 lines, without ANY technical words, you can do this easily for any technical topic really.

Thanks,You are absolutlly right.
I do not have a clear understanding of it's application.
I do not understand how a neural system is applied eigther;,But,I do see the potentual.
If I have a way of working with something;,There is a chance of learning and descovery.
This is a new concept to learn to work with logic relitivity {how logics relate to each other} even if they make no sense to each other. ie; apples are cars. if I say on a chart;,Then I have a clue to the sense.
I'm 66yrs and reached my limit of exploration of this,with my skills.
I think it's something worth passing along {that's all} This allowed me to pass what I can.
I see it being worked into a A.I. system and improving it's logic capibalities,to the point of reasoning.
Accumulating clues and formulating data from logics to direct a neural system process,in a logical way.
I can only show what it is,does, and one possible way of how it fits into a system.
The main thing, it is a function, to come up with a relitive code. {a truth}-- 0.172322 is "info of "apples are cars" and "charts" is at truth0.172322 .txt
This is a clue because truth0.172322 .txt can be a directory of different charts.More clues are needed to pin which one down to see why apples are cars.
A. logical aid to a search engine.
Another potentual:
* = gains
- = losses
He came home with the car and a few flowers.
************************-------------------------********
He came home with flowers.
potentual memorization: {relation}
2*6=12 cd,total 2*6 cd,loss =12 cd,gain function memorized(A$,B$)
3*6=18 total= sub coder(A$)
4*6=24 loss= sub coder(B$)
gain=total+(loss*-1)
{memory}
select case gain
case 0.323344
ans$ "=4"
Ans=4
case 0.006422
ans$ "=12"
Ans=12
ect
This is the best I've got to show what to look for.
{Just for show}
I include my best attempt at seeing it in an neural system,which clearly I do not understand.
But it does poorly demonstrate potentual tie _in to AI.
programed in Just Basic 101
'@@@@@
'TnkStrmStatDcd$ Top_Of_Page
'@@@@@
'
' Ao Ai
Aoc=5:NumAo=Aoc:last=400
Trg$="1"
'Correct$="1010110010001001" '@@@
Letter$="A1@.txt I1.txt E1@.txt O1.txt A2@.txt B1@.txt C1.txt D1.txt E1@.txt I1.txt O1.txt C1.txt B1@.txt D1.txt U1.txt Y1@.txt"
for j=1 to Lng(Letter$)
Filename$=word$(Letter$,j)
if mid$(right$(Filename$,5),1,1)="@" then
Correct$=Correct$;"1"
else
Correct$=Correct$;"0"
end if
next j
dim NeurlWt$(NumAo+1)
'2:168 Min:Sec 2:168 Min:Sec 2:48 Min:Sec
'AoWt$= "0.1734792e-91 -0.27483515e-91 -0.49977805e-91 -0.85970463e-91 -0.15695738e-91"
if AoWt$="" then gosub [IniOWt]
print AoWt$
NeuralWt$="":Ao$=""
for Inter=1 to 7
Esum=0
startTime = time$("seconds")
for Letr=1 to Lng(Letter$)
if Ao$<>"" then AoWt$=Ao$
if Letr>1 then Trg$=""
X2Wt$=""
for Ao=1 to Aoc
NeuralCd$=""
scan
gosub [Clrall]
Path$="C:\AIWarehouse\AISet\Vowels"
Filename$=word$(Letter$,Letr)
FullPath$=Path$;"\";Filename$
open FullPath$ for input as #1
if eof(#1) < 0 then [FskipIt]
index=0
[Floop]
scan
index=index+1
line input #1, Stream$
if Stream$="" then [FskipIt]
gosub [codeExe]
NeuralCd$=NeuralCd$;X1Cd$:X1Cd$=""
NeuralWt$=NeuralWt$;LstWt$:LstWt$=""
if eof(#1) = 0 then [Floop]
[FskipIt]
if Ao=1 then
NeuralCd$=NeuralCd$;"1"
NeuralWt$=NeuralWt$;str$((rnd(-.6)+1e-15)+((rnd(-.3)+1e-15)*-1))
'if Trg$="1" then print NeuralCd$:print NeuralWt$
end if
close #1
'print SumCd
X2Wt$=X2Wt$;str$(Sigmoid(SumCd,Sigmoid));" "
next Ao
Acu=0:Err=0
for Ao=1 to Lng(AoWt$)
scan
Acu=Acu+(val(word$(X2Wt$,Ao))*val(word$(AoWt$,Ao)))
next Ao
Dy=Sigmoid(Acu,Sigmoid)
Df=abs(Dy)
correct=0
correct=val(mid$(Correct$,Letr,1)):Err=(correct+(Df*-1))
if Dy<0 then Df=Err*-1:print tab(16);Err
AccuSum=SumCd:TstSum=SumCd+Err
if Err<>-1 then gosub [LogicActivator]
if correct=1 then Err=Err*5:print tab(16);Err;" *5";" ";"NeurlActivation Letter ";Letr
'print NeuralCd$:print NeuralWt$:print X2Wt$:print AoWt$
if correct=1 then gosub [AdjustWts] ' logic--if present_correct but_not_all then adjust weights
print tab(6);AoWt$
print tab(12);X2Wt$
if correct=1 then Esum=Esum+Err^2
next Letr
print Esum
if Esum=0 then [Found]
next Inter
[Found]
endTime=time$("seconds")
CapTM=(endTime-startTime)
Min=int(CapTM/60):Sec=(CapTM mod 60)
print "process time ";Min;":";Sec;" Min:Sec"
print Correct$
end
'@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ Primary Code processing rutine @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
'@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ Line Entery
'*********************************************************** Coder initialize ******************************************************
[codeExe]
Pi=atn(1)*4:Mdeg=(180/Pi):Mrdn=(Pi/180)
AL$="":TL$=""
GapCd$="":GapTTL$="":FrgALcap$="":FrgTLcap$=""
cnt=0
'********************** execute Coder
'StrmWt$=""
gosub [coder]
return
'*********************************************************** Coder to Cd ***********************************************************
[coder]
T=asc(mid$(Stream$,1,1))
A=45
index=0
[codeLoop]
scan
index=index+1
if asc(mid$(Stream$,index,1))=0 then [CdExit]
D=val(str$(asc(mid$(Stream$,index,1)))) 'NOTE read Asc
'gosub [PotentualCharProcessor]
R=val(str$((D+(1/T))))
c=abs(val(str$((1/(R*360)))))
L=abs(val(str$(((A*(1/360))))))
F=val(str$((1/(((L*c)^.5)*(2*Pi)))))
A=val(str$(atn((1/((R*(1/F))+1)))*Mdeg))
T=val(str$((R*tan(A*Mrdn))))
Rcd=((1/T)+(1/A))
if Trg$="1" then Wt=((rnd(-.6)+1e-15)+((rnd(-.3)+1e-15)*-1))*1e-5:LstWt$=LstWt$;str$(Wt);" "
X1Cd$=X1Cd$;str$(Rcd);" "
SumCd=SumCd+(Rcd*Wt)
AccuWrd=AccuWrd+((1/T)+(1/A))
AccuFrg=AccuFrg+((1/T)+(1/A))
AccuStrm=AccuStrm+((1/T)+(1/A))
AccuPgg=AccuPgg+((1/T)+(1/A))
'gosub [PotentualRcdprocessor]
AL$=AL$+str$(A)+" "
TL$=TL$+str$(T)+" "
if JmpFrg=1 then [FrgJ]
FrgALcap$=FrgALcap$+str$(A)+" "
FrgTLcap$=FrgTLcap$+str$(T)+" "
[FrgJ]
Skip=0
goto [codeLoop]
[CdExit]
Stream$="" 'reset Stream$
return
'@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ Primary DeCode processing rutine @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
'************************************************************** Decoded ***********************************************************
Function DeCoder$(TL$,AL$,byref DL$)
[DeCode]
DL$=""
Pi=(atn(1)*4):Mdeg=(180/Pi):Mrdn=(Pi/180)
indx=0
[DCodeLoop]
indx=indx+1
if word$(TL$,indx)="" then [DCdExit]
T=val(str$(val(word$(TL$,indx))))
A=val(word$(AL$,indx))
DL$=DL$+chr$(int(T*(1/tan(A*Mrdn))))
goto [DCodeLoop]
[DCdExit]
DeCoder$=DL$
end function
'********************************************
[PotentualCharProcessor] '{Char FLAG detection only--Not Words}
JmpFrg=0
if D=126 then print "Fire Flaged"
if D=94 then
AccuFrg=0:JmpFrg=1 'reset AccuFrg - set JmpFrg to flag perse frgment {this} frg_T and frg_A
'###################
FrgALcap$="" 'reset FrgALcap$
FrgTLcap$="" 'reset FrgTLcap$
end if
return
'********************************************************* Coder Procerrors ********************************************************
'********************************************
[PotentualRcdprocessor]
Wrdn=4 '{search perimitor}
if D=32 then '*** Word monitor detector ***
cnt=cnt+1 '{search opperator}
if cnt<>Wrdn then WrdLoss=WrdLoss+AccuWrd
WrdCd$=WrdCd$;str$(AccuWrd);" " 'Cd_capture
WrdTTL=WrdTTL+AccuWrd 'Total_capture
AccuWrd=0 'reset AccuWrd
end if
'*************************
Frgn=2 '{search perimitor}
if D=37 then '*** Fragment monitor detector ***
fgcnt=fgcnt+1 '{search opperator}
if fgcnt<>Frgn then FrgLoss=FrgLoss+AccuFrg
FrgCd$=FrgCd$;str$(AccuFrg);" " 'Cd_capture
FrgTTL=FrgTTL+AccuFrg 'Total_capture
AccuFrg=0 'reset AccuFrg
'###################
print tab(15);DeCoder$(FrgTLcap$,FrgALcap$,DL$) '*** display Fragment Stream$
FrgALcap$="" 'reset FrgALcap$
FrgTLcap$="" 'reset FrgTLcap$
end if
'*************************
return
'@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ ClearAll
[Clrall]
Stream$="":AL$="":TL$="":DL$=""
AccuWrd=0:AccuFrg=0:AccuStrm=0:AccuPgg=0:WrdLoss=0:FrgLoss=0:SumCd=0
X1Cd$="":NeuralCd$="":LstWt$="":FrgALcap$="":FrgTLcap$=""
WrdCd$="":FrgCd$=""
WrdTTL=0:FrgTTL=0:WrdLoss=0
JmpFrg=0:AccuWrd=0:AccuFrg=0:AccuStrm=0:AccuPgg=0
':AoWt$="":X2Wt$="":NeuralWt$=""
return
'*********************************************
[IniOWt]
for j=1 to NumAo
AoWt$=AoWt$;str$((((rnd(-.6)+1e-15)+((rnd(-.3)+1e-15)*-1))*1e-3));" "
next j
return
'*********** accquire String word length ***************
'This function is use to get a string element_count
'Note in this case an element is anything between the spaces of a string
'
Function Lng(byref List$)
index=0
[LngLoop]
index=index+1
Ele$=Word$(List$,index)
if Ele$="" then [LngExit]
goto [LngLoop]
[LngExit]
Lng=index-1
end function
'*********************************************
Function Sigmoid(byref arg,byref Sigmoid)
if arg>1 then arg=300*(1/arg)
sign=1
scan
if arg>-9e-15 and arg<9e-15 then arg=0
if arg<0 then sign=-1
Sigmoid=((Exp((abs(arg)*-1))^2))*sign
'print abs((Sigmoid*100)+(13.5335283*-1))
if abs((Sigmoid*100)+(13.5335283*-1))<1e-7 then Sigmoid=1
end function
'*****************************************************
[LogicActivator]
Fr=0
if TstSum<>0 then Y=abs(-1+(((AccuSum*1e-5)+(TstSum*1e-5))*((1/(TstSum*1e-5))*.5)))
gosub [EntrSig] ':print tab(35); Sigmoid;" ";Y;
if Sigmoid>=(100+(1e-15*-1)) then
Print " Fire";" ";"LgcActivation Letr ";Letr;" ";AcuE:Fr=1
Capture$=Capture$;str$(Letr);" "
end if
return
'********************
[EntrSig]
if abs(Y)>=0 and abs(Y)<=(1+1e-15) then Sigmoid=Sigmoid(Y,Sigmoid)*100
return
'*********************************************
[AdjustWts]
Mu=0.5e-5
SlopeA=Slope(Y)
'print tab(22); SlopeA
Acu=0
for indxAo=1 to Lng(AoWt$) ' X2Cnt
Acu=val(word$(X2Wt$,indxAo))
SlopeX=Slope(Acu)
for indxNx=1 to Lng(NeurlCd$) 'X1Cnt
scan
if SlopeX<>0 then Corrshn=val(word$(NeurlCd$,indxNx))*SlopeX*val(word$(AoWt$,indxAw))*SlopeA
'print tab(30);Corrshn
for indxj=1 to Lng(NeurlCd$) 'X1Cnt
LtrW=val(word$(NeuralWt$,indxj))
HfHWt$=HfHWt$;str$(val(word$(LtrW$,indxj))+Corrshn*Err*Mu);" "
next indxj
NeuralWt$=HfHWt$
next indxNx
next indxAo
'*************************
WAh$=AoWt$:AoWt$=""
for indxAo=1 to Lng(WAh$) ' X2Cnt
Corrshn=val(word$(X2Wt$,indxAo))*SlopeA
'print tab(32);Corrshn
AoWt$=AoWt$;str$(((val(word$(WAh$,indxAo))+Corrshn*Err*Mu)*1e-8)*.7);" "
next indxAo
'print tab(45);AoWt$
'print
return
'***********************************************
function Slope(byref Acu)
Slope=(Acu*(1+(Acu*-1)))
end function
.

This topic is closed to new replies.

Advertisement