5.3 Partitioned Model
This topic describes the information about the partitioned model in Machine Learning.
Oracle in-Database machine learning allows the user to design partitioned models.
Partitioned model organizes and represents multiple models as partitions in a single model entity, enabling the user to easily build and manage models tailored to independent slices of data.
Table 5-7 Example - Customer details
CUSTOMER ID | BRANCH | MARITAL STATUS | INCOME | SAVING | LUXURY SPEND |
---|---|---|---|---|---|
12345 | BRT | Y | 15000 | 6723 | 1000 |
12346 | BRT | N | 17500 | 250 | 750 |
12347 | CSR | D | 25000 | 2654 | 1900 |
12348 | GRF | N | 16567 | 20 | 2500 |
In this above example of data, the user could build a single partitioned model on independent slices of data based on branch code.
The user has the advantage of having a single partitioned model instead of having multiple models for each individual branch.
Parent topic: Machine Learning Framework