Creating A Framework for Validating AI Models by Harnessing the Power of Synthetic Data

Time: 1:01 pm
day: Pre-Conference Day


The role of synthetic data in pressure testing systems and validating AI models has emerged as a crucial area of research and development. To fully utilize its potential, it is necessary to explore the significance of synthetic data and its ability to measure the sensitivity of AI systems to specific data types.. By drawing from recent work and practical experiences, we propose a comprehensive framework for leveraging synthetic data in the validation of AI models. By addressing the challenges posed by black box systems and undefined data, this framework aims to provide a standardized approach to testing and validation.

Join this workshop to:

  • Explore the complexities of working with uncertain data sets
  • Examine the implications of not knowing the presence or absence of signals
  • Leverage synthetic data offers a well-defined environment where signals and noise are precisely delineated
  • Navigate the uncharted territory of AI validation techniques and present a roadmap for evaluating the sensitivity of different approaches.