Report Portal

An MDX Outlier Solution

Anomalies in data can skew analysis leading to incorrect conclusions and subsequent actions that could be useless or even damaging as it will address the wrong problem. Anomalies in data are usually attributed to human data entry error or perhaps are the result of complex ETL calculations that may be absurdly wrong (like a NaN value or ?2,147,483,648, the -1 INT value). Even though manual measurements of things such as vital signs during a physical exam are becoming a thing of the past (using electronic devices linked directly to a computer), devices used to take measurements are still prone to the occasional anomaly. An anomaly could represent readings under very unusual circumstances or it could represent a bug in the device.


Tags: design, mdx


2007-2015 VidasSoft Systems Inc.