Research:
Foundations of the
Cybernetics of Goal-orientated Systems
Cybernetics of Goal-orientated Systems
Starting point for our basic research are simple feedback systems. They are the simplest goal-orientated systems. And they show already some fundamental capabilities:
- They show basic cognitive abilities, namely the comparison of sensor data and goal-values.
- They apply these comparisons as inputs in decisions.
- And they are the simplest structures, which can make elementary decisions. These elementary decisions have the general form:
if {(data) (relation) (goal-value)}, then {trigger for a goal-orientated action}
More complex goal-orientated systems build on the structures of feedback systems. So they result from enlarging feedback systems.
And this enlarging shows a necessary, parallel evolution: Increasingly complex controller structures enable enhanced cognitive capabilities and more complex decision-making. And here every later stage of this evolution presupposes the previous ones, along the developmental line feedback systems - enlarged feedback systems - adaptive systems - systems with individual pattern recognition - with individual sequence learning - anticipatory systems, etc.
We see the investigation of the interrelation of controller structures, cognitive capabilities and abilities for decision-making as the base for a cybernetic epistemology. Furthermore it provides the background for the general analysis of goal-orientated behavior (see applications).
Below we offer links to basic papers. If you do not have access via these links, or have any questions or comments, please write us a mail - we answer shortly: office@nechansky.co.at
Details:
- Cybernetics as science of decision making:
An introduction to the structural analysis of simple controllers, their cognitive functions and their capabilities for making elementary decisions. Shows, too, how controller structure ar integrated in living systems. .
External Link:
Nechansky (2011a), Cybernetics as science of decision making
- Basic Aspects of Epistemology and Cybernetics:
This brief chapter written for the 'Handbook of Human Computation' provides an introduction to basic questions of epistemology and social epistemology and their relation to the cybernetics of decisions; it contains additionally a short overview on the epistemlogical options offered by computers.
External Link:
Nechansky (2013d), Epistemological Issues in Human Computation
- Feedback Systems as starting point:
This basic paper on the cybernetics of goal-orientated systms shows how feedback systems make elementary decisions. And it proposes, too, a cybernetic definition of information.
- Enlarging feedback systems:
Here we make a first step towards complex goal-orientated systems: The enlargement of a feedback system by adding a simple internal strcuture that can recognize changes.
Download:
Nechansky (2008c), Functional and Structural Requirements for Goal-orientated Systems
- The route towards complex goal-orientated systems:
The paper analyzes systematically how feedback systems can be enlarged, by adding one ore more functional elements like sensors, deciders, memory, etc. This dry investigation shows major structural constrains and design rules for the way from feedback systems to complex controllers, including, as we suggest, brains.
External Link:
Nechansky (2009a), Design Rules for Complex Goal-orientated Systems
- Preprogrammed adative systems:
Preprogrammed adative systems can make some situation-specific decisions to select given behavioral options. This can happen in hierarchical or one-level structures. The paper suggests that the (seldom explored) one-level adaptive systems are the base for the evolution towards the brain.
External Link:
Nechansky (2010b), Preprogrammed Adapitve Systems
- Adaptive Systems that can develop system-specific behavior:
Adaptive systems, which can develop individual behavior, must be able (1) to make a decision to deviate from default behavior, (2) to trigger any new one, and (3) to evaluate the sucess of this new behavior in relation to the hierarchical highest, existential goal-values of the system. Here trial-and-error emerges as a cybernetic necessity.
External Link:
Nechansky (2010d), Adaptive Systems that Can Develop System-specific Behavior
- Learning systems:
The paper analyzes how a goal-orientated system can store observations as new standards for pattern matching and can apply them later on for pattern recognition. Developing new behavior towards such newly recognized patterns requires the trial-and-error mechanism of adaptive systems, determining finally the effect of new patterns on the highest, existential goal-values of the systems. Here emerges an internal 'emotional' or 'psychic' evaluation of external patterns as a cybernetic necessity. So the paper determines the cybernetic starting point of individual psychology.
External Link:
Nechansky (2012b), Pattern Recognition, Learning and the Base of Individual Psychology
- Sequence learning systems:
The paper analyzes how sequence learning can build on pattern recognition. This requires additional structures which register repeatedly occurring patterns and connect these to sequences. Possiblities for this connection of patterns are discussed. Finally a simple 'associative' way using directed and hierarchical connections is analyzed in detail. It shows basic features of neural nets.
External Link:
Nechansky (2012a), Sequence Learning Systems
- Output-side attention directing systems:
Complex subsystems for pattern recognition and sequence learning may recognize more than one pattern or sequence at a point in time; so they can deliver simultaneously various output signals. Therefore they require an output-side attention directing system to decide, which of these signals should be used to determine the further behavior of the whole system.
External Link:
Nechansky (2012c), Output-side Attention Directing Systems
- Elementary Anticipatory Systems:
Elementary anticipation, understood as anticipation of the reoccurance of one known pattern, can emerge out of sequence learning. It requires (1) the identification of the beginning of a known sequence and (2) using one later element of this sequence as an anticipated pattern. Additionally it requires a subsystem for switching between feedback (pattern recognition) and feedforward (anticipation).
External Link:
Nechansky (2013a), Elementary Anticipatory Systems
- Complex Anticipatory Systems:
Complex anticipation, understood as anticipation of the reoccurance of a known sequence of patterns, builds on elementary anticipation. It requires the same functions to identify the beginning of known sequences, but needs complex additional structures to search and decide for anticipating a possible continuation of such a sequence. An up - down search processes, as found in the neural nets of the brain, is suggested as mechanism to handle that.
External Link:
Nechansky (2013b), Complex Anticipatory Systems
This series on the cybernetics of goal-orientated systems and cybernetic epistemology will be continued.
© 2005 - 2024 Dr. Helmut Nechansky