Fractal cognitive computing node, computer-implemented method for learning procedures, computational cognition cluster and computational cognition architecture

A computing device for learning procedures, said computing device comprising a first fractal cognitive computing node (FCN) the FCN comprising:

• a first input, configured to receive a first input signal, said first input signal comprising a first set of pairs <key, value>, a key of the first set of pairs representing an identifier and a value of the first set of pairs representing a magnitude of the first input signal;
• a first output, configured to provide a first output signal, said first output signal comprising a second set of pairs <key, value>, a key of the second set of pairs representing an identifier and a value of the second set of pairs representing a magnitude of the first output signal;
• a second input, configured to receive a second input signal, said second input signal being a first output signal of another FCN and comprising a third set of pairs <key, value>, the key of the third set of pairs representing an identifier and the value of the third set of pairs representing a magnitude of the second input signal;
• a third input, configured to receive a spatial attention (SA) parameter comprising a single dimension value;
• a memory, configured to store a collection of items, each item having a set of tuples <key, value>, and each item representing a previously stored input from at least one of said first or second input signals;
• and
• a processing unit comprising one or more processors, configured to implement a computational function that:
• compares the first input signal or the second input signal or a combination of said first and second input signals to some or all of the stored collection of items;
• and
• calculates a similarity measure for each compared item;
• wherein said first output signal outputted via said first output being calculated as a selection of the compared items having a similarity measure greater than said SA parameter;
• wherein when the selection of compared items is empty, a new item is added to said memory and the set of tuples <key, value> of the new item and the first output signal are set according to the first input signal, to the second input signal or to a combination of the first and second input signals;
• wherein when the selection of compared items is not empty, the first output signal is set to said selection;
• wherein the first fractal FCN is configured to be replicated into a new FCN by replicating at least one of the items stored in the memory of the first FCN in the new FCN and by connecting a first output of the new FCN to the first input of the first FCN and a second input of the new FCN to the first output of the first FCN,
• wherein the new FCN is replicated from the first FCN based on at least on one of the items stored in the memory of the first FCN;
• the first new output of the new FCN is coupled to the first input of the first FCN;
• the second new input of the new FCN is coupled to the first output of the first FCN,
• wherein at least two FCNs connected to each other form a Computational Cognition Cluster having a hierarchical structure, wherein the Computational Cognition Cluster further comprises
• a first input, configured to receive a first input signal from an output of a children Computational Cognition Cluster, said first input signal comprising a compositional signal of tuples;
• an output, configured to connect the Computational Cognition Cluster to a parent Computational Cognition Cluster via a first input of the parent Computational Cognition Cluster;
• and
• a second input, configured to receive a second input signal from an output of the parent Computational Cognition Cluster;
• and
• a third input, configured to receive a conditioning (CD) parameter comprising a single dimension value.