oracle - Integrating Classification Algorithms in OBIEE 11g? -
currently, i've been involved in data warehouse based banking analytics project. we've been using oracle database , oracle business intelligence tool (obiee 11g) front-end or presentation layer user through weblogic server.
here, need implement support vector machine (svm) algorithm atm card fraud pattern detection & cart churn prediction , feed result user through obiee. question how can integrate these algorithms (implemented in python) obiee?
i've seen oracle data miner integration obiee implements own algorithm.
there new feature in 11g external tables called preprocessor
(white papaer 11gr2 can found here) external table enables read flat files directly tables in oracle. new preprocessos command enables specify script name instead of file , oracle use output stream input. way can algorithem results directly oracle.
lets take cart algorithm example
create working directory in oracle
create or replace directory dmdir "/some_folder_on_your_os"
create script executes algorithm , prints output in comma delimited format. lets cart algorithm can output tree
level,attribute,value,outcome,confidence
.create external table
create table cart_tree level number, attribute varchar2(2000), value varchar2(2000), outcome varchar2(2000), confidence number ) organization external ( type oracle_loader default directory dmdir access parameters ( records delimited newline fields terminated ',' missing field values null preprocessor execdir:'my_cart_algorithem.py' fields ( level, attribute, value, outcome, confidence
)) location ('')) reject limit unlimited
- notice
location
parameter script. in case non.
- notice
query
select * cart_tree
it's important understand every time query table oracle executes script. - external tables create materialized view on top of them.
a second option might find useful using existing integration oracle has r. both cart , svm algorithm available in r , has build in integration oracle. can read here here
Comments
Post a Comment