Chao Jen Cheng

US Patent:

20080319679, Dec 25, 2008

Filed:

Oct 3, 2007

Appl. No.:

11/866964

Inventors:

Lei M. Li - Los Angeles CA, US
Chao Cheng - Los Angeles CA, US
Huanying Ge - Los Angeles CA, US

Assignee:

University of Southern California - Los Angeles CA

International Classification:

G06F 17/18
G01N 33/48

US Classification:

702 20, 702179

Abstract:

The present invention discloses methods and systems for analyzing microarray data. The method includes the general steps of providing microarray data, normalizing the data using a least trimmed squares regression, and then analyzing the normalized microarray data to obtain a desired result such as an expression profile. There is also disclosed a method of subdividing an array into subarrays before normalization. This approach provides a method for improving measurement accuracy and salvaging array data from arrays containing minor defects. Also disclosed is a Probe-Treatment-Reference (PTR) model for streamlining normalization and summarization of microarray data by allowing multiple references. Other aspects of the present invention include computer systems and computer readable media encoding methods of the present invention.

US Patent:

20130151164, Jun 13, 2013

Filed:

Nov 19, 2012

Appl. No.:

13/681129

Inventors:

University of Southern California - Los Angeles CA, US
Chao CHENG - Los Angeles CA, US
Huanying GE - Los Angeles CA, US

Assignee:

UNIVERSITY OF SOUTHERN CALIFORNIA - Los Angeles CA

International Classification:

G06F 19/20

US Classification:

702 19, 702179

Abstract:

The present invention discloses methods and systems for analyzing microarray data. The method includes the general steps of providing microarray data, normalizing the data using a least trimmed squares regression, and then analyzing the normalized microarray data to obtain a desired result such as an expression profile. There is also disclosed a method of subdividing an array into subarrays before normalization. This approach provides a method for improving measurement accuracy and salvaging array data from arrays containing minor defects. Also disclosed is a Probe-Treatment-Reference (PTR) model for streamlining normalization and summarization of microarray data by allowing multiple references. Other aspects of the present invention include computer systems and computer readable media encoding methods of the present invention.

US Patent:

20210357730, Nov 18, 2021

Filed:

May 12, 2020

Appl. No.:

16/872979

Inventors:

- George Town, KY
Chao CHENG - San Mateo CA, US
Yang JIAO - San Mateo CA, US

International Classification:

G06N 3/04
G06F 17/18
G06F 17/15

Abstract:

Systems and methods for improved convolutional layers for neural networks are disclosed. An improved convolutional layer can obtain at least two input feature maps of differing channel sizes. The improved convolutional layer can generate an output feature map for each one of the at least two input feature maps. Each input feature map can be applied to a convolutional sub-layer to generate an intermediate feature map. For each intermediate feature map, versions of the remaining intermediate feature maps can be resized to match the channel size of the intermediate feature map. For each intermediate feature map, an output feature map can be generated by combining the intermediate feature map and the corresponding resized versions of the remaining intermediate feature maps.

US Patent:

20170265210, Sep 14, 2017

Filed:

Mar 10, 2017

Appl. No.:

15/456128

Inventors:

- San Diego CA, US
Srinivas KATAR - Fremont CA, US
Hao ZHU - Milpitas CA, US
James CHO - Mountain View CA, US
Alireza RAISSINIA - Monte Sereno CA, US
Chao ZOU - San Jose CA, US
Sabita NAHATA - Newark CA, US
Guido Robert FREDERIKS - Watsonville CA, US
Raja BANERJEA - San Jose CA, US
Chao CHENG - Santa Clara CA, US
Ning ZHANG - Saratoga CA, US
Zhanfeng JIA - Belmont CA, US

International Classification:

H04W 72/12
H04L 5/00
H04W 72/04

Abstract:

Apparatuses and methods for separately scheduling and grouping multiple wireless local area network (WLAN) users are disclosed. The apparatuses and methods include scheduling, at an access point (AP), multiple wireless stations (STAs) associated with the AP; identifying based at least in part on the ranking, a group of STAs from the STAs for transmission of data over a single orthogonal frequency-division multiple access (OFDMA) frame in a WLAN; and providing information resulting from scheduling the STAs and identifying the group of STAs to at least one of the STAs. In others aspects, the apparatuses and methods include receiving at an STA from an AP, scheduling information for a group of STAs including the STA; and transmitting based on the scheduling information, data over an OFDMA frame in an unlicensed or shared spectrum, the STA having a traffic load different from the traffic load of the at least one additional STA.