Wednesday Oct 08, 2014

2014 was a very good year for Oracle Advanced Analytics at Oracle Open World 2014

2014 was a very good year for Oracle Advanced Analytics at Oracle Open World 2014.   We had a number of customer, partner and Oracle talks that focused on the Oracle Advanced Analytics Database Option.    See below with links to presentations.  Check back later to OOW Sessions Content Catalog as not all presentations have been uploaded yet.  :-(

Big Data and Predictive Analytics: Fiserv Data Mining Case Study [CON8631]

Moving data mining algorithms to run as native data mining SQL functions eliminates data movement, automates knowledge discovery, and accelerates the transformation of large-scale data to actionable insights from days/weeks to minutes/hours. In this session, Fiserv, a leading global provider of electronic commerce systems for the financial services industry, shares best practices for turning in-database predictive models into actionable policies and illustrates the use of Oracle Data Miner for fraud prevention in online payments. Attendees will learn how businesses that implement predictive analytics in their production processes significantly improve profitability and maximize their ROI.

Developing Relevant Dining Visits with Oracle Advanced Analytics at Olive Garden [CON2898]

Olive Garden, traditionally managing its 830 restaurants nationally, transitioned to a localized approach with the help of predictive analytics. Using k-means clustering and logistic classification algorithms, it divided its stores into five behavioral segments. The analysis leveraged Oracle SQL Developer 4.0 and Oracle R Enterprise 1.3 to evaluate 115 million transactions in just 5 percent the time required by the company’s BI tool. While saving both time and money by making it possible to develop the solution internally, this analysis has informed Olive Garden’s latest remodel campaign and continues to uncover millions in profits by optimizing pricing and menu assortment. This session illustrates how Oracle Advanced Analytics solutions directly affect the bottom line.

A Perfect Storm: Oracle Big Data Science for Enterprise R and SAS Users [CON8331]

With the advent of R and a rich ecosystem of users and developers, a myriad of bloggers, and thousands of packages with functionality ranging from social network analysis and spatial data analysis to empirical finance and phylogenetics, use of R is on a steep uptrend. With new R tools from Oracle, including Oracle R Enterprise, Oracle R Distribution, and Oracle R Advanced Analytics for Hadoop, users can scale and integrate R for their enterprise big data needs. Come to this session to learn about Oracle’s R technologies and what data scientists from smart companies around the world are doing with R.

Extending the Power of In-Database Analytics with Oracle Big Data Appliance [CON2452]

The need for speed could not be greater—not speed of processing but time to market. The problem is driven by the long journey data takes before evolving into insight. Insight, however, is always relative to assumption. In fact, analytics is often seen as a battle between assumption and data. Assumptions can be classified into three types: related to distributions, ratios, and relations. In this session, you will see how the most-valuable business insights can come in the matter of hours, not months, when assumptions are challenged with data. This is made possible by the integration of Oracle Big Data Appliance, enabling transparent access to in-database analytics from the data warehouse and avoiding the traditional long journey of data to insight.

Market Basket Analysis at Dunkin’ Brands [CON6545]

With almost 120 years of franchising experience, Dunkin’ Brands owns two of the world’s most recognized, beloved franchises: Dunkin’ Donuts and Baskin-Robbins. This session describes a market basket analysis solution built from scratch on the Oracle Advanced Analytics platform at Dunkin’ Brands. This solution enables Dunkin’ to look at product affinity and a host of associated sales metrics with a view to improving promotional effectiveness and cross-sell/up-sell to increase customer loyalty. The presentation discusses the business value achieved and technical challenges faced in scaling the solution to Dunkin’ Brands’ transaction volumes, including engineered systems (Oracle Exadata) hardware and parallel processing at the core of the implementation.

Predictive Analytics with Oracle Data Mining [CON8596]

This session presents three case studies related to predictive analytics with the Oracle Data Mining feature of Oracle Advanced Analytics. Service contracts cancellation avoidance with Oracle Data Mining is about predicting the contracts at risk of cancellation at least nine months in advance. Predicting hardware opportunities that have a high likelihood of being won means identifying such opportunities at least four months in advance to provide visibility into suppliers of required materials. Finally, predicting cloud customer churn involves identifying the customers that are not as likely to renew subscriptions as others.

SQL Is the Best Development Language for Big Data [CON7439]

SQL has a long and storied history. From the early 1980s till today, data processing has been dominated by this language. It has changed and evolved greatly over time, gaining features such as analytic windowing functions, model clauses, and row-pattern matching. This session explores what's new in SQL and Oracle Database for exploiting big data. You'll see how to use SQL to efficiently and effectively process data that is not stored directly in Oracle Database.

Advanced Predictive Analytics for Database Developers on Oracle [CON7977]

Traditional database applications use SQL queries to filter, aggregate, and summarize data. This is called descriptive analytics. The next level is predictive analytics, where hidden patterns are discovered to answer questions that give unique insights that cannot be derived with descriptive analytics. Businesses are increasingly using machine learning techniques to perform predictive analytics, which helps them better understand past data, predict future trends, and enable better decision-making. This session discusses how to use machine learning algorithms such as regression, classification, and clustering to solve a few selected business use cases.

What Are They Thinking? With Oracle Application Express and Oracle Data Miner [UGF2861]

Have you ever wanted to add some data science to your Oracle Application Express applications? This session shows you how you can combine predictive analytics from Oracle Data Miner into your Oracle Application Express application to monitor sentiment analysis. Using Oracle Data Miner features, you can build data mining models of your data and apply them to your new data. The presentation uses Twitter feeds from conference events to demonstrate how this data can be fed into your Oracle Application Express application and how you can monitor sentiment with the native SQL and PL/SQL functions of Oracle Data Miner. Oracle Application Express comes with several graphical techniques, and the presentation uses them to create a sentiment dashboard.

Transforming Customer Experience with Big Data and Predictive Analytics [CON8148]

Delivering a high-quality customer experience is essential for long-term profitability and customer retention in the communications industry. Although service providers own a wealth of customer data within their systems, the sheer volume and complexity of the data structures inhibit their ability to extract the full value of the information. To change this situation, service providers are increasingly turning to a new generation of business intelligence tools. This session begins by discussing the key market challenges for business analytics and continues by exploring Oracle’s approach to meeting these challenges, including the use of predictive analytics, big data, and social network analytics.

There are a few others where Oracle Advanced Analytics is included e.g. Retail GBU, Big Data Strategy, etc. but they are typically more broadly focused.  If you search the Content Catalog for “Advanced Analytics” etc. you can find other related presentations that involve OAA.

Hope this helps.  Enjoy!


Monday Jan 18, 2010

Fraud and Anomaly Detection Made Simple

Here is a quick and simple application for fraud and anomaly detection.  To replicate this on your own computer, download and install the Oracle Database 11g Release 1 or 2.  (See for more information).  This small application uses the Automatic Data Preparation (ADP) feature that we added in Oracle Data Mining 11g.  Click here to download the CLAIMS data table.  [Download the .7z file and save it somwhere, unzip to a .csv file and then use SQL Developer data import wizard to import the claims.csv file into a table in the Oracle Database.]

First, we instantiate the ODM settings table to override the defaults.  The default value for Classification data mining function is to use our Naive Bayes algorithm, but since this is a different problem, looking for anomalous records amongst a larger data population, we want to change that to SUPPORT_VECTOR_MACHINES.  Also, as the 1-Class SVM does not rely on a Target field, we have to change that parameter to "null".  See for detailed Documentation on ODM's anomaly detection.

drop table CLAIMS_SET;

exec dbms_data_mining.drop_model('CLAIMSMODEL');

create table CLAIMS_SET (setting_name varchar2(30), setting_value varchar2(4000));


insert into CLAIMS_SET values ('PREP_AUTO','ON');


Then, we run the dbms_data_mining.create_model function and let the in-database Oracle Data Mining algorithm run through the data, find patterns and relationships within the CLAIMS data, and infer a CLAIMS data mining model from the data.  


dbms_data_mining.create_model('CLAIMSMODEL', 'CLASSIFICATION',




After that, we can use the CLAIMS data mining model to "score" all customer auto insurance policies, sort them by our prediction_probability and select the top 5 most unusual claims.  

-- Top 5 most suspicious fraud policy holder claims

select * from

(select POLICYNUMBER, round(prob_fraud*100,2) percent_fraud,

rank() over (order by prob_fraud desc) rnk from

(select POLICYNUMBER, prediction_probability(CLAIMSMODEL, '0' using *) prob_fraud


where PASTNUMBEROFCLAIMS in ('2 to 4', 'more than 4')))

where rnk <= 5

order by percent_fraud desc;

Leave these results inside the database and you can create powerful dashboards using Oracle Business Intelligence EE (or any reporting or dashboard tool that can query the Oracle Database) that multiple ODM's probability of the record being anomalous times (x) the dollar amount of the claim, and then use stoplight color coding (red, orange, yellow) to flag only the more suspicious claims.  Very automated, very easy, and all inside the Oracle Database!

Everything about Oracle Data Mining, a component of the Oracle Advanced Analytics Option - News, Technical Information, Opinions, Tips & Tricks. All in One Place


« November 2015