Data quality assessment
DataRobot will also perform a data quality assessment and notify you if it finds any data issues, as shown in the following screenshot:
Figure 5.2 – Data quality issues
In this case, it has found outliers in eight features. You can look into the details to see if these look acceptable or if you need to drop or otherwise fix these outliers. We will do this as we explore and analyze each of these features in the following section.
Notice that it also looked for any disguised missing values or excess zeros in any feature. These can be hard to detect manually and can be problematic for your models, so it is important to fix these issues if they come up. For example, you saw in Chapter 4, Preparing Data for DataRobot, that we already fixed the issue of excess zeros in the normalized-losses feature. If we had not done that previously, DataRobot would alert us to fix this or filter out those rows before proceeding. It will also perform...
