Integrating LSS Principles, Tools, and Techniques in Quality 4.0
Concurrent Sessions (M01 - M06)
Presentation Description: In the rapidly changing world of analytics most data are collected over time, yet a time component is often ignored in the analysis of data. As a Six Sigma practitioner you are likely familiar with analyzing data from designed experiments where any time component is handled by good randomization, but what happens with observational data occurring over time? In this presentation, we will move one step further in your analytics journey by understanding the impact of time on your analysis. Using case studies based on recent customer engagements, I will discuss the benefits and challenges with using time in your models. We will begin by looking at common patterns in time series data, including autocorrelation, trend and seasonality. By understanding these patterns, you will learn how to create simple forecasting models using time as the only predictor and then consider strategies for handling time as one of many potential predictors. It’s about time to start thinking about time!