User:Mathiastck/P==NP

I thought about 3D level design

I realized a lot of the time it's a mistake.

Any 3D Map can be converted to a 2D Map after all, and it is good to reduce a harder problem to an easier problem.

Take your 3D Map.

Construct a connected 3d graph with no negative edges.

Now, using a brute force divide an conquer method, find the largest planar, IE, the largest 2D planar connected graph which you can construct from subsets of our 3D graph.

Make this your new working 2D map.

Now pick 3 connected vertices on our original 3d map. Using a brute force divide and conquer method construct new versions of our original 3d map where these 3 vertices are replaced with a single vertex, that is connected to every other vertex connected to any of the above 3 vertices.

Take your working 2D map. Using a brute force divide and conquer method, begin one at a time adding vertices to our 2D map.

Keep track of the state of every working 2D map constructed.

Prove that this works for a 3D map with one vertex.

Prove it works for a 3D map with 2.

Prove that it if it works for a 3D map with X vertices it will work for a 3D map with X+1 vertices.

Now claim,

this is your major claim. Call it claim A.

YOU CAN REDUCE ANY 3D MAP TO A 2D MAP IN POLYNOMIAL TIME.

Claim B.

Any NP problem can be converted to a P problem in constant time. It will be solved in polynomial time X where X is a constant describing the difficulty of problem P or problem NP as I am proving that,

CLAIM C

P==NP.

Claim D,

here is another way to convert an NP problem to a P problem.

Assume for the purposes of contradiction that P == NP, and C is = the complexity of a given NP problem A1, and a given P problem B1, where C1 = the complexity of problem A1 and C2 = the complexity of B1.

Now, for C = 0.

Return true.

For C = 1. Solve A1, in polynomial time, on a single Turing Machine T sub C. Where C = 1. Call this turing machine C1.

Prove that in this case, C1 = C2.

P1 can be reduced to P2, and vice versa.

NOW FOR CASE C sub x, Cx.

Run X turing machines, one a time. Alternating, giving each machien an equal amount of time, on average, but giving each machine at least 1/2 of the averatge time per machine and giving no machine more the 2x the average time per machine.

Let each machine return its results one at a time.

Now for a given class of problems, defined as NAVIGATION PROBLEMS, OR MAPREDUCE PROBLEMS, OR SEARCH PROBLEMS, OR SORTING PROBLEMS.

I believe I have reduced to NP to P.

And I want my money.

Check plz.

QED

Send all checks to Matt Kanninen, in San Jose.

I work for Sony.

All rights reserved, with the exception of those rights which can released under an Apache or BSD style license.

I also license this under the creative commons license.

I also license this under GNU iff, this does not conflict with any previous statements.

All author rights are mine, Matt Kanninen, matt.

Mathiastck (talk) 16:38, 31 January 2010 (UTC)