General Electric Company

Paul A Delano

Michael James Hartman

Richard I Hartley

Kenneth B Welles, Ii

David W Tong

Abhijit Chatterjee

An artificial neural network detects points in feature space outside of a boundary determined by a set of sample data. The network is trained using pseudo data which compensates for the lack of original data representing 'abnormal' or novel combinations of features. The training process is done iteratively using a net bias parameter to close the boundary around the sample data. When the neural net stabilizes, the training process is complete. Pseudo data is chosen using several disclosed methods.

Method for using a feed forward neural network to perform classification

An auxiliary monolithic integrated circuit chip provides both buffer amplification and testing interfaces. Off-the-shelf monolithic integrated circuit chips can be connected into an electronics system using one of these auxiliary buffer chips before each input port and after each output port, to implement functional testing similar to that done on monolithic integrated circuit chips with built-in test circuitry.

Buffer integrated circuit providing testing interface

Built-in test circuitry, which is appropriate for monolithic integrated circuit chips that are to be connected in a plural-chip package, uses electronic token passing to select one of the test input ports in the circuitry to be tested for application of test input vectors. The built-in test circuitry also uses electronic token passing to select one of the test output ports in the circuitry to be tested from which test results are to be supplied. Methods for testing based on these token passing procedures are described.

Growing community of inventors

Clifton Park, NY, United States of America

Paul A Delano