In an increasingly competitive marketplace, evidence-based decision making has become a requirement for manufacturers. But even with an abundance of manufacturing data on hand, the necessary evidence may not be so easy to tease out.
In this paper, author Tony Cooper argues that the most successful analytical frameworks today combine both statistics and what he terms "smart machine learning." To achieve this balance, Cooper explains, manufacturers need to build both process knowledge – understanding the unique aspects of the data you collect – and personnel knowledge – understanding the perspective of data scientists who do the collecting. Hence, he has developed a three-pronged approach whereby practitioners consider machine learning alongside statisticians and causal discovery.
With its thought-provoking examples of the unique opportunities and challenges of modern manufacturing, this white paper outlines essential best practices for the kind of exploratory data tasks that drive smarter machine learning.