Procurement Summary
Country : Netherlands
Summary : Verification and Validation Methods for Machine Learning Algorithms
Deadline : 13 Aug 2024
Other Information
Notice Type : Tender
TOT Ref.No.: 104172450
Document Ref. No. : 1-12349
Competition : ICB
Financier : Other Funding Agencies
Purchaser Ownership : Public
Tender Value : Refer Document
Purchaser's Detail
Name :Login to see tender_details
Address : Login to see tender_details
Email : Login to see tender_details
Login to see detailsTender Details
Tenders are invited for Verification and Validation Methods for Machine Learning Algorithms - Expro+.
define verification and validation methods for applications relying on Machine LearningtechniquesDescriptionCurrently, Machine Learning (ML) is starting to be integrated in more and more areas of the spaceindustry, some of them are safety or mission-critical (e.g. vision-based navigation). Onefundamental challenge for this adoption process is the limited ability of verifying the results of MLalgorithms and establishing the required confidence in them. In particular, for supervised learningalgorithms, at runtime there is no base truth available for the validation, so conventional validationmethods prove to be ineffective.An effort has to be made to assess the available formal and non-formal verification and validationmethods in this context and define a subset of guidelines and recommendations for specific usecases and types of machine learning.This activity encompasses the following tasks:- Analysis of the ML techniques intended to be used in the frame of critical applications.- Analysis of verification and validation techniques that can be applied on the identified MLtechniques- Identification of representative use cases- Selection or development of the ML models related to the use cases- Application of the verification and validation techniques- Assessment and recommendationsSoftware shall be delivered under ESA Software Community Licence.
Open Date: 11/07/2024 17:02 CET
Closing Date: 13/08/2024 13:00 CET
Price Range: 200-500 KEURO
Documents
Tender Notice