Michael W Lewicki

US Patent:

6799170, Sep 28, 2004

Filed:

Jul 22, 2002

Appl. No.:

10/202190

Inventors:

Michael S. Lewicki - Pittsburgh PA
Terrence J. Sejnowski - Solana Beach CA

Assignee:

The Salk Institute for Biological Studies - La Jolla CA

International Classification:

G06N 302

US Classification:

706 20, 706 22, 706 25

Abstract:

A computer-implemented method and apparatus that adapts class parameters, classifies data and separates sources configured in one of multiple classes whose parameters (i. e. characteristics) are initially unknown. A mixture model is used in which the observed data is categorized into two or more mutually exclusive classes. The class parameters for each of the classes are adapted to a data set in an adaptation algorithm in which class parameters including mixing matrices and bias vectors are adapted. Each data vector is assigned to one of the learned mutually exclusive classes. The adaptation and classification algorithms can be utilized in a wide variety of applications such as speech processing, image processing, medical data processing, satellite data processing, antenna array reception, and information retrieval systems.

US Patent:

20050149462, Jul 7, 2005

Filed:

Sep 24, 2004

Appl. No.:

10/949710

Inventors:

Michael Lewicki - Pittsburgh PA, US
Terrence Sejnowski - Solano Beach CA, US

Assignee:

The Salk Institute for Biological Studies - La Jolla CA

International Classification:

G06E001/00
G06E003/00
G06G007/00
G06F015/18
H03F001/26
G06F015/00
H04B015/00

US Classification:

706020000, 702190000

Abstract:

A computer-implemented method and apparatus that adapts class parameters, classifies data and separates sources configured in one of multiple classes whose parameters (i.e. characteristics) are initially unknown. A mixture model is used in which the observed data is categorized into two or more mutually exclusive classes. The class parameters for each of the classes are adapted to a data set in an adaptation algorithm in which class parameters including mixing matrices and bias vectors are adapted. Each data vector is assigned to one of the learned mutually exclusive classes. The adaptation and classification algorithms can be utilized in a wide variety of applications such as speech processing, image processing, medical data processing, satellite data processing, antenna array reception, and information retrieval systems.

US Patent:

6424960, Jul 23, 2002

Filed:

Oct 14, 1999

Appl. No.:

09/418099

Inventors:

Michael S. Lewicki - Pittsburgh PA
Terrence J. Sejnowski - Solana Beach CA

Assignee:

The Salk Institute for Biological Studies - La Jolla CA

International Classification:

G06N 302

US Classification:

706 20, 600310, 600515

Abstract:

A computer-implemented method and apparatus that adapts class parameters, classifies data and separates sources configured in one of multiple classes whose parameters (i. e. characteristics) are initially unknown. The data set may be generated in a dynamic environment where the sources provide signals that are mixed, and the mixing parameters change without notice and in an unknown manner. A mixture model is used in which the observed data is categorized into two or more mutually exclusive classes. The class parameters for each of the classes are adapted to a data set in an adaptation algorithm in which class parameters including mixing matrices and bias vectors are adapted. Each data vector is assigned to one of the learned mutually exclusive classes. In some embodiments the class parameters may have been previously learned, and the system is used to classify the data and if desired to separate the sources. The adaptation and classification algorithms can be utilized in a wide variety of applications such as speech processing, image processing, medical data processing, satellite data processing, antenna array reception, and information retrieval systems.