11g database logo

New Version - Oracle Spreadsheet Add-In for Predictive Analytics


The Oracle Spreadsheet Add-In for Predictive Analytics (PA) enables Microsoft Excel users to mine their Oracle or Excel data using three simple, " one click" predictive analytics features that automate data mining Using PA, more users can harness the power of Oracle Data Mining without requiring them to possess the knowledge of advanced data analysis. The Profile, Predict and Explain PAs require the Oracle Database Enterprise Edition (EE) Database and the Oracle Data Mining Option.

Key Features:

Explain PA makes it easy for users to better understand their data. Microsoft Excel users simply connect to an Oracle Database, either selects a table or view, or point to data in an open Excel spreadsheet, select a target attribute, for example "High_Value_Customers", and press OK. Explain PA sifts through the data searching for attributes that have the greatest influence on a target attribute, ranks them in order, and displays the results in Excel. For example, Explain PA could be used to identify attributes most associated with:

  • Customers who churn
  • Patients who develop a disease
  • Parts that fail

Profile PA finds customer segments and their profiles, given a target attribute.  Profile PA generates multiple profiles for each target value that can provide greater insight into the data, relationships and customer segments.  For example, one possible profile of customers with Churn_Risk = "HIGH" may be described as:

"AGE < 35" and "INCOME > $45,000" and "OUT_OF_PLAN_MINS >  "$75.".

Predict PA enables users to make intelligent predictions based on data and hidden patterns relationships found by Oracle Data Mining. For each unique identifier e.g. Customer_ID, Predict PA automatically intuits a predictive data mining model from the data and applies that model to data to make predictions. Predict PA makes predictions about categorical attributes e.g. 0 or 1 and can also predict a numeric value.  Predict PA provides a confidence probability for each prediction. Predict PA could be used to predict:

  • Which customers are likely to purchase a product
  • Who is likely to be a “high lifetime value” customer
  • The value of a home

Import/Export Microsoft Excel data quickly and easily to your Oracle Database. Simply connect to the database, highlight the spreadsheet data and the Spreadsheet Add-In for Predictive Analytics creates a new Database table or appends the data to an existing Database table. Similarly, export data from an Oracle Database table to your spreadsheet.

Application developers can use Oracle Data Mining's DBMS_PREDICTIVE_ANALYTICS PL/SQL packages to build advanced business intelligence applications that add new insights and actionable business intelligence to any report or dashboard. For example, a current “ Top Customers” report could run predictive analytics PL/SQL packages inside the Oracle Database and display the “ Predicted Top Customers” report to the user in the report or dashboard for more proactive business intelligence.

 Oracle 11.1 Predictive Analytics Spreadsheet Add-In February 2008
Use Predictive_Analytics.xla for Excel 2003 and earlier
Use Predictive_Analytics_2007.xla for new features with Excel 2007
The 2007 xla is not compatible with with earlier versions of Excel. The xla for older versions of Excel functions with Excel 2007 in compatibly mode; however, the new Excel 2007 features are not available.

The latest version of the Add-In works with both the Oracle Data Mining option to Oracle 10gR2 (10.2) Enterprise Edition and the Oracle Data Mining option to Oracle 11gR1 (11.1) Enterprise Edition. If you intend to use Profile, you must connect to an Oracle 11.1 Enterprise Edition database with the Oracle Data Mining option.

 Oracle 10.2 Predictive Analytics Spreadsheet Add-In February 2006
 Oracle 10.1 Predictive Analytics Spreadsheet Add-In March 2005
Sample Screen Image