Explore a 22-minute conference talk from the Uncertainty in Artificial Intelligence (UAI) 2023 conference focusing on "Parity Calibration" in sequential regression settings. Delve into the concept of parity calibration, which addresses the challenge of predicting whether future observations will increase or decrease compared to current ones. Learn how the researchers demonstrate that extracting parity probabilities from forecasted distributions can lead to poor performance and theoretical unpredictability. Discover how the team leverages the connection between parity calibration and binary calibration to develop an effective online binary calibration method. Examine the practical applications of this approach through real-world case studies in epidemiology, weather forecasting, and model-based control in nuclear fusion. Access the presentation slides to gain a visual understanding of the research findings and methodologies discussed in this insightful talk.