View Full Version : Trinary Temporal Node Network

5th-April-2014, 02:37 AM
The Network
All nodes have the same number of connections.
Any node can connect to any other node but not in self reference.
A group of nodes is selected to be the value of input.
A separate group of nodes is selected to be the value of output.

Changing States
A node has three states (1, -1, 0)
A node is 1 or -1 when its threshold is reached but 0 when not reached.
A node has 2 thresholds.
There is a time step interval of 2 sets of connections for both thresholds.
One set of connections is referred to at a time for the node to be in its current state.
Threshold is reached when adding the states of all connections exceeds 0.

Changing Connections
Each connection is assigned a weight not used in calculating threshold.
Each weight is strengthened by 1 when its node is not 0 and weakens by -1 when it is 0.
When a node is not 0 it changes its connections based of their weights.
Half of the connections with the weakest weights change and the new connections start at weight 0.

Self Organization
Over time loops form of meta-stabilization when reacting to temporal input.
Time Hierarchies forms in abstract layering of loops as connections change.
By evolving the network can be grouped with other self similar networks.
To reduce entropy nodes in the network attract inputs that increase maximum duration of not 0 state.

5th-April-2014, 01:24 PM
I'm too stupid and daft to understand. Actually, could you explain to me what this is about?

5th-April-2014, 01:37 PM
Sounds like AI but I have no idea how it works.

Hey Ak are you high?

5th-April-2014, 02:01 PM
Out of curiosity, where should I start if I want to get into AI? My resources are limited and I am still in highschool.

5th-April-2014, 03:00 PM
Discuss it with me, indeed I'll make some threads about it.