Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn to develop explainable forecasting models through causal inference in this 25-minute conference talk from MLOps World. Discover a novel causal model that directly performs predictions without requiring additional forecasting models, moving beyond traditional approaches that only use causal inference for feature selection. Explore the Xi-Correlation (XiCorr) method, a new nonparametric statistic based on data ranking, combined with Dynamic Conditional Correlation (DCC) to analyze time-varying causality among variables. Master the technique of forecasting target variables through polynomial regression using immediate causal predictors identified by XiCorr and validated by DCC methods. Understand how causal scores provide clear explanations of each predictor's effect on the forecast. Examine practical applications through a Canadian macroeconomic indexes dataset case study that demonstrates promising backtest results, bridging the gap between machine learning, operations research, and explainable AI in production environments.
