Share via


Predictor Schema Example

This example shows all the related tables and entries for a model configuration named PurchaseCfg1.

PredictorModelCfgs

ModelCfgName SiteName
PurchaseCfg1 Retail

This example assumes that the PurchaseCfg1 model configuration consists of three sources of data:

Demographic information from the User table (additional rows not shown).

UserID Age Gender Education
jilluser@roguecellars.com 23 F 19
barneyuser@arborshoes.com 49 M 3
jackuser@frogkick.com 28 M 13

Product Purchase information from the Purchases table (additional rows not shown).

UserID SKU QTY
jilluser@roguecellars.com wine_fine_au_zin_19 3
jilluser@roguecellars.com tv_big_36_m3 1
jilluser@roguecellars.com pumps_bl_6 1

Ad Click information from the AdClicks table (additional rows not shown).

UserID URL Count
jilluser@roguecellars.com www.thewinecellar.com/ad_zin12 3
jilluser@roguecellars.com www.thewinecellar.com/ad_cab3 2
jilluser@roguecellars.com www.arborshoes.com/ad4c 1

PredictorDataTables

ModelCfgName TableName Type Case
Column
Pivot
Column
Aggregate
Column
Aggregate
Type
PurchaseCfg1 User 0 UserID <NULL> <NULL> <NULL>
PurchaseCfg1 Purchases 1 UserID SKU QTY 0
PurchaseCfg1 AdClicks 1 UserID URL Count 0
PurchaseCfg1 PurchaseCfg1_Attributes 2 Not applicable <NULL> <NULL> <NULL>

The entries in the following table override the default behavior. The first three entries construct a hierarchy where "tv_big_36_m3" is the child of "TVs" and "TVs" is the child of "Electronic Devices". "Electronic Devices" stands at the top of the hierarchy and therefore does not have a parent. The fourth entry specifies all of the SKUs to have a Discrete and Modeled As Binary distribution. The fifth row specifies to use the Age attribute to predict other attributes, but not to be predicted. This means that no Decision Tree predicting the Age attribute will be built. The entry also specifies the Distribution to be Continuous, Lognormal and not Modeled As Binary.

PredictorAttributes_PurchaseCfg1

Prop
ID
Parent
ID
Name Table
Name
Column
Name
Distri-
bution
UseTo
Predict
Predict
1 0 Electronic Devices <NULL> <NULL> <NULL> <NULL> <NULL>
2 1 TVs <NULL> <NULL> <NULL> <NULL> <NULL>
3 2 tv_big_36_m3 <NULL> <NULL> 4 True True
4 0 SKU <NULL> <NULL> 4 True True
5 0 Age <NULL> <NULL> 2 True False

The following table shows two models built from the PurchaseCfg1 model configuration. The first model, Purchase1, was built as a prediction (type 0) model. The second model, Demog1, was built as a segment (type 1) model.

PredictorModels

Model
Name
ModelCfg
Name
Model
Type
Date
Created
Build
Time
Measured
Accuracy
Sample
Fraction
Measured
Accuracy
Max
Predictions
K
Purchase1 PurchaseCfg1 0 getdate() 30000 .05 10 <NULL>
Demog1 PurchaseCfg1 1 getdate() 45000 .05 10 10

(PredictorModels continued)

Max
Buffer
Size
Input
Attribute
Fraction
Output
Attribute
Fraction
Sample
Size
Complexity
Penality
Minimum
CasesTo
Split
Data
Fit
Score
Recommend
Score
Data
<NULL> 1 1 -1 100 5 11.08 0.3946 <Binary>
1000000 1 1 16000 <NULL> <NULL> 4.177 0.1874 <Binary>


All rights reserved.