China
Africa
Explore chapters and articles related to this topic
Lean – The Relentless Pursuit to Eliminate Waste
Published in Raymond Kelly, The Myths and Truths of Lean Transformations, 2018
Some examples of over-processing include: having a process that’s more complexed than required, e.g. automating a process that can be completed at the same or better cost and quality manually, repeating a step of a process just to make sure, etc.;producing something to tighter specifications than required by the customers;inspecting something to a tighter tolerance or Acceptable Quality Limit (or Level) than required;process steps that do not add value to the customer and are not wanted by the customer;carbon copy (i.e. “cc”) many people on an email instead to only relevant recipients; hitting “reply all” to an email instead of sending only to relevant recipients.
Modeling and Change Detection for Count-Weighted Multilayer Networks
Published in Technometrics, 2020
Hang Dong, Nan Chen, Kaibo Wang
In real-world applications, the interactions are sometimes directional and include other quantitative information. For example, there are senders and recipients for phone calls, text messages, and E-mails indicating the contact direction from sender to recipient; the activities in online social networks such as “mention” and “following” are also directional. When these interactions are aggregated at some specific time scale, we get a directed weighted network with the edge weights representing the frequency or intensity of the interactions. For example, Figure 1 shows a two-layer network of E-mails among 13 employees of Enron Corporation. The first layer represents the ordinarily sent E-mails, and the second layer represents the carbon copy/blind carbon copy (cc/bcc) E-mails. Each edge represents the E-mails from the sender to the receiver using cc/bcc, with the weight (expressed through edge width in the figure) representing the frequency of such E-mails. It is obvious that the interaction patterns of these two layers are similar but have layer-specific features as well. It is clear from this figure that the directed weighted version of a social network contains richer information than the undirected binary representation (Gao et al. 2017), and could help us better understand the underlying behavior of a network. Therefore, multilayer network models are required to capture the layer-specific information as well as the correlations among different layers for this kind of interaction data.