Diving Deeper

Abstract:The LEARN process is an alternative statistical method that delivers valuable information about a process without the need for heavy duty math. By following its stages--look, examine, assign, regression and natural selection (LEARN)--one can compile a rich set of input data to deliver valuable outputs. Subject matter experts bring system engineering expertise to an initial statistical review to understand relevant data and processes and provide a rough sketch of the process flow. Knowledge gained in the 'look' phase is used to improve the model being developed, refine variables, and produce a plot. Data is progressively structured and added prior to the 'regression' phase. Through natural selection, differences between data and the developing model are diminished, resulting in a continuously improving …

Access this article
Other ways to access this article

Social Bookmarking

Digg, delicious, NewsVine, Furl, Google, StumbleUpon, BlogMarks, Facebook



I really like the concept of this article. The quality is lacking.

In trying to fully understand the concept and method being described, I tried to replicate the analysis performed. What I found is that the data represented as the X value in the “Adjusted Subsystem” column of Table 1 is not supported by the “Adjustment Applied” values.

Additionally the regression equation and r-squared shown in Figure 6 is not supported by the data shown in the far right two columns of Table 1 – labeled as “Plotted data (x, y)”.

Bummer.

--Scott Crowley, 10-14-2016



--RM, 10-04-2016


Featured advertisers