---

Methods_and_Procedures_for_the_Verification_and_Validation_of_Artificial_Neural_Networks_-_Brian_J



Preface
Artificial neural networks are a form of artificial intelligence that have
the capability of learning, growing, and adapting within dynamic
environments. With the ability to learn and adapt, artificial neural networks
introduce new potential solutions and approaches to some of the more
challenging problems that the United States faces as it pursues the vision of
space exploration. For instance, one of the areas of potential problems is in
air and ground vehicle control. To be successful on manned missions to
Mars and the Moon, intelligent adaptive systems, such as neural networks,
may be needed to assist in crew operations to accommodate an ever
changing environment. The major obstacle to deploying such highly
complex systems is the verification and validation of these systems.
The need is being recognized by organizations such as NASA, that the
supporting function of verification and validation must be brought to bear
for neural network systems to gain the necessary acceptance within their
respective problem domains. As the facility responsible for assuring
software safety, reliability, and quality of programs and missions, the NASA
Independent Verification and Validation (IV&V) Facility will be
increasingly challenged to certify and evaluate software systems that contain
neural network technologies. The NASA IV&V Facility has recognized the
need and importance of neural network technology as it is becoming more
feasible for use in future space applications. To address this need, the
NASA rV&V Facility sponsored the Institute for Scientific Research, Inc.
(ISR) under Research Grant NAGS-12069 through the NASA Goddard
Space Flight Center, to research and develop methodologies for the
independent verification and validation of artificial neural networks.
This book is a result of three years of research conducted for the NASA
IV&V Facility and examines some of the more promising methods and
procedures for the verification and validation of artificial neural networks
and adaptive systems. This book does not endorse artificial neural networks
as the perfect solution, but instead disseminates the methods and procedures
for verifying and validating these highly complex systems so that they can
be used in safety-critical and mission-critical applications. The methods and
procedures presented in this book were chosen because of their applicability,
technology maturity level, technical feasibility, and usability in the
verification and validation of neural networks.

İndir:
http://rapidshare.com/files/105614281/M...n_J._.html

---

Başa dön