We have discussed about how to create a model in Einstein Discovery in our previous article. As analysis and training progress, Einstein Discovery scrutinizes your data for quality concerns, including duplicate impact (known as the multicollinearity data alert), potential biases, frequent missing values, and various other data quality issues. Whenever a potential data quality problem arises, Einstein promptly alerts you with a data alert.
The Data Alerts panel shows you each occurrence and gives you the option to either take action or ignore the alert.
In our model, Einstein detected multicollinearity in the data. The problem is that two or more variables (Lead Source and 5 other variables) are highly correlated to each other which could have a duplicate impact on the outcome.
Take appropriate action and hit Next.
Since you have addressed the data alerts, Einstein will create a new model version. Give a description to the new version.
Einstein re-analyzes and re-trains the model in a new version. When the new version is complete, you’ll see the model performance overview again. In the new version, you see the new version number and there’s no longer alerts to be reviewed.
The Area Under the Curve (AUC) represents the rate of correct classification by a logistic model. An AUC of 0.5 means that the model performs no better than random guessing. An AUC of 1.0 means that the model correctly classifies data 100% of the time, which can indicate data leakage.
The threshold value tells your model how to classify a binary outcome based on the calculated probability. The default threshold is 0.5, but you can move this value up or down to accommodate your use case.
The distribution of the outcome variable (in this case, ‘Converted’) can be viewed as a histogram for the binary values ‘0’ and ‘1’, where the later indicates the total number of converted leads in the lead data set.
There will be a list of top predictors( input variables) which found relevant to the predicted outcome. They will be listed in the order of their percentage of importance in prediction.
From the ‘Data Insights’ section, you can anlayze the data in detail.
To learn more, click Introduction to Einstein Discovery (salesforce.com).
Take Five Consulting is a technology company, based in Virginia U.S., that specializes in the Mortgage Banking vertical especially LOS implementation and application development. Take Five Consulting creates and implement mortgage technology and software specifically for Mortgage Industry.
