Algorithmic principle for neural networks.

From: Martin Michael Musatov <musatovatattdotnet_at_gmail.com>
Date: Thu, 12 Jul 2012 13:14:52 -0700 (PDT)
Message-ID: <01a477f0-0d9f-4eca-9be2-5ddd48862092_at_d24g2000yqh.googlegroups.com>



Consider networks based on pattern recognition as having vision. Define vision as pattern recognition. The network has the ability to recognize relevant to by user interaction with results. Users interact with relevant results. A high amount of relevant results in a period of time is defined by a lower amount of interaction in equal time. Based on the time-hierarchy all machines are programmed to accept the following statement as the first true statement found to be always true in all cases future and past having been checked and verified as true to both machine and human verifiers goes good in the future equals went well in the past after this result is recorded and checked and verified as true the truth of this statement is found and this statement labels eternal truth. For the purposes of producing a result all statements are either false or true. All false statements are always false in all cases past present or future and a true statement is define as a statement found to be true in at least one instance and all true statements are further defined as being either true or eternally true. Eternally true statements are all true statements verified as eternally true in all instances past present and future and a true statement is simply a statement true in one instance. All eternally true statements are true and all true statements are true and some true statements are eternally true. Received on Thu Jul 12 2012 - 22:14:52 CEST

Original text of this message