Marcin Szeliga - Explainable and responsible Artificial Intelligence

🔹Explainable and responsible Artificial Intelligence
🔹Marcin Szeliga
Artificial Intelligence is one of the fastest-growing areas of IT for all businesses. Models are expected to make more and more autonomous decisions. But sometimes they fail and we want to know why. Unfortunately, even Data Scientists often don’t understand how their models behave. To build responsible AI systems assessing the model’s performance is not enough. We need to incorporate explainability into AI systems to understand the decisions they make and ensure they align with ethical and legal guidelines. Explainability allows us to trace the decision-making process of AI models and identify any biases or errors that may have occurred. In this session, we will discuss various techniques used to achieve explainability in AI models available on Azure Machine Learning Services and SynapseML.