JMP in the Upstream 06: What's my hydrocarbon? Using spectral data to predict octane rating
Created:
Original session date: 3 February 2022
Topics covered: Data import and cleaning, graphing spectral lines, Generalized Regression, and PLS
In this episode of JMP in the Upstream, @Bill_Worley (Sr. Systems Engineer) showcases the add-in he co-developed for analyzing spectral data in JMP. He walks us through methods for importing the data into JMP, data cleaning and preparation, tips for visualizing the data prior to analysis, and the analysis itself. He demonstrated several ways to analyze the data including Partial Least Squares (PLS) and Generalized Regression. In the latter method he is able to reduce the number of spectra required to draw actionable decisions from 400 (total lines collected), to 340 (required for a statistically significant PLS model), to less than 70 (Gen Reg). Imagine only needing to collect <20% of the data you originally thought? That is real time and real money saved!
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.