Models - Teradata Vantage

Machine Learning Engine Analytic Function Reference

Product
Teradata Vantage
Release Number
9.02
9.01
2.0
1.3
Published
February 2022
Language
English (United States)
Last Update
2022-02-10
dita:mapPath
rnn1580259159235.ditamap
dita:ditavalPath
ybt1582220416951.ditaval
dita:id
B700-4003
lifecycle
previous
Product Category
Teradata Vantageā„¢

To use the Attribution_MLE function, you set up either one or two models. Each model has two parts:

Model Part Description
Attribution model Specifies which event types are eligible for attribution.
Distribution model Specifies how attribution is distributed across eligible events.

Attribution Model Types

Attribution Model Type Description
SIMPLE All events are eligible for attribution.
EVENT_REGULAR Specifies event types eligible for attribution.
EVENT_OPTIONAL Specifies event types eligible for attribution only if no other event is eligible.
SEGMENT_ROWS Specifies number of rows prior to conversion event in which events are eligible for attribution.
SEGMENT_SECONDS Specifies number of seconds prior to conversion event in which events are eligible for attribution.

How Many Models You Need

In the following situations, you must set up two models:

Situation First Model Second Model
You want to specify the window using both rows and seconds. SEGMENT_ROWS SEGMENT_SECONDS
You want to specify optional events. EVENT_REGULAR EVENT_OPTIONAL

In other situations, you must set up only one model.

Distribution Model Types

Distribution Model Type Description
UNIFORM Attribution for conversion event is divided evenly across all preceding qualifying events.
LAST_CLICK Attribution for conversion event is assigned entirely to most recent qualifying event.
FIRST_CLICK Attribution for conversion event is assigned entirely to least recent qualifying event.
WEIGHTED Attribution for conversion event is assigned to qualifying events with weights you specify.
EXPONENTIAL Attribution for conversion event is assigned to qualifying events from most recent to least recent with exponential decay, using decay factor and time unit you specify.