Organizations are often concerned with sharing of sensitive data internally (and definitely externally). The existing data privacy techniques, however, are counter-productive to the data science effort. Emerging techniques such as synthetic data, differential privacy and federated learning provide a best-of-both-worlds approach. This session provides in-depth insights into these techniques.

The use of predictive models creates trust challenges for enterprises. Model bias and lack of transparency are primary challenges for the trust gap. This session will demystify the approaches to mitigate bias and making the models explainable leveraging organizational people, improved processes and platform enhancements.