What are counted as Complex Features in DataVisor platform?

Complex Features are features of complex calculation logics and are computationally heavy. The following 4 types of features are defined as Complex Features when counted towards your license usage:

1. Velocity feature: a velocity feature can be identified by having a "Aggregation" type in the feature type column in the DataVisor console. In a velocity feature, you are using aggregation logics to conduct a velocity computation (e.g. Count, Sum, Max, Average, Distinct_Count) of a dimension (e.g. transaction_amount) against another dimension (e.g. user_ID) over a time window (e.g. Past 7 days) 

2. Graph feature: a graph feature is type of complicated velocity features that involve 3 or more than 3 dimensions in the aggregation calculations, e.g. "Dintinct_count of Emails used by the User_IDs who used a specific IP_address in the Past 7 days". Graph feature will also show a "Aggregation" type in the Feature type column.

3. Regular Expression feature: a Regular Expression (Regex) feature is a feature that uses the regex() operator to conduct a sequence of symbols and characters expressing a string or pattern to be searched for within a longer piece of text. 

4. Data Source feature: a Data Source feature queries an external data source (e.g. a third party data provider) or a reference list stored on Datavisor's platform in real time to retrieve additional data for data enrichment. 

5. Model feature: a feature that executes a machine learning model in its script. Model feature is often also referred to as model score feature. 

Was this article helpful?
1 out of 1 found this helpful



Please sign in to leave a comment.