[This is preliminary documentation and is subject to change.]

This class implements the Non Action Centered Subsystem (NACS) in the Clarion Library

The NonActionCenteredSubsystem..::..NonActionCenteredSubsystemParameters type exposes the following members.

Properties

  Name Description
Public property APPRAISAL_ITERATION_COUNT
Specifies the number of reasoning iterations to perform when doing appraisals
Public property BOTTOM_LEVEL_WEIGHT
The weight by which bottom level activations are applied following a reasoning iteration
Public property CONCLUSION_THRESHOLD
The threshold conclusions must pass to be inferred
Public property DEFAULT_PARTIAL_MATCH_OFF_THRESHOLD
The threshold value when partial match is turned OFF (used for knowledge retrieval)
Public property DEFAULT_PARTIAL_MATCH_ON_THRESHOLD
The threshold value when partial match is turned ON (used for knowledge retrieval)
Public property K
The number of conclusions to be returned (when KWTA is used)
Public property KWTA_SELECTION_TEMPERATURE
The temperature to use for stochastically selecting conclusions (when KWTA is used)
Public property OFFLINE_LEARNING_FREQUENCY
The frequency in which a check is made to see whether "offline" learning should occur
Public property PARTIAL_MATCH_ON
Turns partial match on/off (used for knowledge retrieval)
Public property PARTIAL_MATCH_THRESHOLD
The threshold a chunk strength must pass for the chunk to be matched
Public property PERFORM_APPRAISAL
Specifies whether the NACS should perform appraisals
Public property REASONING_ITERATION_COUNT
The number of reasoning iterations to perform
Public property REASONING_ITERATION_TIME
The time (in milliseconds) that it takes to perform a reasoning iteration
Public property RESET_BEFORE_REASONING_ROUND
Indicates whether the current state of the NACS's reasoning mechanism (as set by the previous round of reasoning) should be go before a new round of reasoning is started
Public property RETURN_INPUTS_AS_CONCLUSIONS
Specifies whether chunks, included as part of the input into the NACS's reasoning mechanism, should be returned as part of the conclusions from reasoning
Public property RULE_DELETION_FREQUENCY
The frequency in which associative rule deletion should occur
Public property TOP_LEVEL_WEIGHT
The weight by which top level activations are applied following a reasoning iteration
Public property USE_BOTTOM_LEVEL
Specifies whether or not the bottom level of the NACS's reasoning mechanism should be used
Public property USE_KWTA
Specifies whether the "K Winner Take All" selection method should be used before conclusions are returned from reasoning
Public property USE_TOP_LEVEL
Specifies whether or not the top level of the NACS's reasoning mechanism should be used

See Also