Loren Data's SAM Daily™

FBO DAILY ISSUE OF MAY 29, 2011 FBO #3473
DOCUMENT

A -- Investigate Active and Proactive Machine Learning Methods research project for Salt Lake City VA Medical Center. - Attachment

Notice Date

5/27/2011
 

Notice Type

Attachment
 

NAICS

541519 — Other Computer Related Services
 

Contracting Office

Department of Veterans Affairs;Rocky Mountain Network;VISN 19 Contracting;4100 E. Mississippi Avenue, Suite 900;Glendale CO 80246
 

ZIP Code

80246
 

Solicitation Number

VA25910RP0265
 

Response Due

6/1/2011
 

Archive Date

7/31/2011
 

Point of Contact

Pamela Barnes
 

Small Business Set-Aside

N/A
 

Description

NOTICE OF INTENT TO SOLE SOURCE: The Department of Veterans Affairs, Veterans Integrated Service Network (VISN) 19 Rocky Mountain Network Contracting Center 4100 East Mississippi Avenue, Suite 900, Glendale, CO 80246 issues this notice of intent to award a contract on a sole source basis on behalf of the George E. Whalen VA Medical Center Salt Lake Research and Development Office, Health Services Research and Development Office, Consortium for Healthcare Informatics Research (CHIR) Program with the Language Technologies Institute (LTI) in the School of Computer Science at Carnegie Melon University to produce active and proactive machine learning methods. This requirement will be procured in accordance with FAR Part 6.302-1, only one responsible source and no other supplies or services will satisfy agency requirements. The Research Service of the Salt Lake City Healthcare System requires specialized services in machine learning for medical informatics research. In the vast majority of inductive learning problems, including centrally induced patterns in clinical data, where there is a paucity of high-quality expert-labeled data. Diagnosis of PTSD requires the careful consideration of a number of factors, including details related to service, symptoms, demographics, and mental condition. These factors are recorded across a number of document types and in a variety of formats including structured, semi-structured and unstructured values. The unique qualifications of the Language Technologies Institute (LTI) is based on its ™ expertise in conducting research and providing graduate education in all aspects of language technologies, including computational linguistics, machine translation, speech recognition and synthesis, statistical language modeling, information retrieval and web search engines, text mining, information management, digital libraries, intelligent tutoring, and more recently bio-sequence/biolanguage, structure and function analysis (genome, proteome). The LTI combines linguistics approaches with machine learning and corpus-based methods, depending on the scientific questions investigated and project needs. The Language Technologies Institute (LTI) has expert experience in providing reports in the investigation of Active and Proactive Machine Learning Methods. The LTI has expertise in providing national reports on the methods of Active learning to address the issue of selecting the most informative instances to give to the expert to label, based on criteria such as: 1) maximal uncertainty of the learning method based on prior training data (ask where you know the least), 2) maximal density unsampled regions in instance space (ask for the most representative new cases), 3) maximal diversity in instance space (ask where no one asked before). The Language Technologies Institute has the expertise on the next steps beyond active learning which is proactive learning, where the contractor has consulted on the aspects of expert performance to determine the cost of expert labeling, the estimated accuracy of expert labeling, and which of multiple external experts should be queried. For instance, some instances may be easier (less-costly) to label by any expert than other more complex instances. In other cases, one expert can be more accurate or more available or less costly than another or there may be tradeoffs. The Language Technologies Institute (LTI) has developed programs utilizing heterogeneous features to improve clinical classification and outcome predictions. The Language Technologies Institute (LTI) has tested the performance of various automated approaches required as an iterative consideration of the contributions of feature sets. The Language Technologies Institute (LTI) has the research experience where they have completed a number of experiments which will be conducted to measure the performance of combinations of features including structured values (e.g., demographics, labs) as well as NLP-derived concepts indexed from free text. THIS NOTICE OF INTENT IS NOT A REQUEST FOR COMPETITIVE RESPONSES. However, all responses received within five days after publication of this notice may be considered by the government. Responses can be faxed to Pamela Barnes at 303-691-6558 or sent via mail to VISN 19 Rocky Mountain Network Contracting Center, 4100 East Mississippi Avenue, Suite 900, Glendale, CO 80246. A determination not to compete the acquisition based upon the responses received is solely at the discretion of the government. Information received will normally be considered solely for the purpose of determining whether to conduct a competitive procurement.
 

Web Link

FBO.gov Permalink
(https://www.fbo.gov/spg/VA/VARMCCC/VARMCCC/VA25910RP0265/listing.html)
 

Document(s)

Attachment
 

File Name: VA-259-10-RP-0265 VA-259-10-RP-0265_2.doc (https://www.vendorportal.ecms.va.gov/FBODocumentServer/DocumentServer.aspx?DocumentId=204949&FileName=VA-259-10-RP-0265-000.doc)

Link: https://www.vendorportal.ecms.va.gov/FBODocumentServer/DocumentServer.aspx?DocumentId=204949&FileName=VA-259-10-RP-0265-000.doc

 

Note: If links are broken, refer to Point of Contact above or contact the FBO Help Desk at 877-472-3779.
 

Place of Performance

Address: Salt Lake City VA Medical Center;500 Foothill Drive;Salt Lake City, Utah

Zip Code: 84148
 

Record

SN02458872-W 20110529/110527234031-1537d179cb9c0aa5c73f2050c8ffc345 (fbodaily.com)
 

Source

FedBizOpps Link to This Notice
(may not be valid after Archive Date)

---

| FSG Index  |  This Issue's Index  |  Today's FBO Daily Index Page |