tag:blogger.com,1999:blog-590043897961646114.post3601296891584946184..comments2024-02-06T02:06:06.364-08:00Comments on Engaging Market Research: Variable Selection in Market Segmentation: Clustering or Biclustering?Unknownnoreply@blogger.comBlogger5125tag:blogger.com,1999:blog-590043897961646114.post-91583836043984193772014-08-02T11:41:41.057-07:002014-08-02T11:41:41.057-07:00I just posted a follow-up that ought to help. Bicl...I just posted a follow-up that ought to help. Biclustering calculates similarity between respondents using only some of the variables. Traditional clustering calculates distances using all the variables. Obviously, it is more complicated than this one point. But if forced to pick one aspect that separates biclustering from clustering, I would focus on biclustering’s attempt to differentiate among individuals using different variables. Put simply, biclustering ignores all the many variables with small discrepancies that get summed together to mislead and concentrates on the few variables with the largest differences. The language is a bit figurative, though it makes the point.Joel Cadwellhttps://www.blogger.com/profile/14946447393733294251noreply@blogger.comtag:blogger.com,1999:blog-590043897961646114.post-47878985615145608182014-08-02T11:21:36.682-07:002014-08-02T11:21:36.682-07:00I just posted an example using a dataset from a R ...I just posted an example using a dataset from a R package that you can access and run your own analyses. The post is called "Customer Segmentation Using Purchase History." I hope it helps.Joel Cadwellhttps://www.blogger.com/profile/14946447393733294251noreply@blogger.comtag:blogger.com,1999:blog-590043897961646114.post-55928356672469383272014-07-31T13:03:55.620-07:002014-07-31T13:03:55.620-07:00Excellent! The point you made amplifies our possib...Excellent! The point you made amplifies our possibilities in tying to capture the complexity of real-world data...thank youAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-590043897961646114.post-80896507373944823612014-07-30T01:18:11.716-07:002014-07-30T01:18:11.716-07:00Nice write-up. From a substantial perspective, is ...Nice write-up. From a substantial perspective, is their a difference to the traditional market segmentation approach, i.e. apply first a factor analysis and then the cluster algorithm? <br />As highlighted by the previous comment, an example illustrating meaningful differences between the two approaches in a real-world scenario would be a great follow-up.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-590043897961646114.post-5884540932191532542014-07-30T00:13:29.730-07:002014-07-30T00:13:29.730-07:00Very insightful! Although we do Segmentation exerc...Very insightful! Although we do Segmentation exercises I've never actually thought of using a Bi-Clust. What would be your opinion in terms of translating a bi-clust model to make real world sense to clients. They already have a hard time understanding a regular clustering exercise! But am looking forward to some example that you would showcase !! Unknownhttps://www.blogger.com/profile/00675716111484535267noreply@blogger.com