Breve NN Abstract Data Type (ADT)

• Nodes
•     int create-input-node with-name thisName = "input-node" (string) with-bias thisBias = 0.0 (float)  with-initial-activation thisActivation =0.0 (float) with-transfer-function  thisTransferFunction = "Logistic" (string)
•     int create-hidden-node ...
•     int create-output-node ...
•    
•     string get-node-name for-node thisNode (int).
•     int set-node-name for-node thisNode (int) to thisString (string).
•     get & set bias
•     get & set initial activation
•     get & set transfer function
•     get-current-activation
•    
•     int delete-node thisNode (int).

• Links
•     int create-link from-node thisNode (int) to-node thatNode (int) with-weight = 0.0 (float).
•     int set-link from-node thisNode (int) to-node thatNode (int) to-weight (float).
•     float get-link from-node thisNode (int) to-node thatNode (int).
•    int connection-exists from-node thisNode (int) to-node thatNode (int).
•    int delete-link from-node thisNode (int) to-node thatNode (int).

• Patterns
•     int add-pattern with-inputs theseInputs (list) with-outputs theseOutputs (list).
•     int delete-pattern at-position thisPosition (int).
•     list get-all-patterns.

• initializing
•     int initialize-network with-init-function thisFunction = "Randomize_Weights" (string) with-init-parameters theseParameters = {-1,1} (list).

• training
•     list train with-function thisFunction = "Std_BackPropagation" (string)  for-epochs thisManytimes = 1 (int) with-parameters theseParameters = {.2} (list).
•     float test.

• running
•     list run with-inputs theseInputs (list).

• files
•     int load-patterns from-file "thisFile".
•     int save-patterns to-file "thisFile".
•     int load-network from-file "thisFile".
•     int save-network to-file "thisFile".