Contact Sales: (619) 330-0780

Predict! Issue #40

Dear Readers,

Those of us in the data science community know that what we do is fairly complex. The average person on the street isn’t very likely to know much about big data analytics, much less be able to develop predictive models to address critical business issues. If this stuff were easy, a data scientist wouldn’t need advanced degrees in statistics, applied mathematics, computer science or other hard sciences.

The challenging aspects of predictive analytics lie in developing algorithms and models that make accurate predictions. This is the core activity that comprises data science, of course. Deploying and using the predictive model can also be challenging in many organizations, distracting data scientists from their core mission, but it doesn’t need to be that way. This is where Zementis eliminates the complexity, so that data scientists can stay focused on what they do best…. data science!

Zementis offers organizations many ways to simplify the deployment of predictive models and accelerate business insights, irrespective of the front-end approaches (data mining tools, predictive modeling techniques) and end-state deployment environment for the models (cloud or on-premise, Hadoop or in-database). This month, we wanted to highlight two ways that make life easier for data scientists and the business users who they serve.

Pentaho’s Data Integration platform, as well as MySQL, the world’s most popular open source database, work seamlessly with Zementis’ ADAPA product. ADAPA simplifies and turbo-charges the process of deploying predictive models, while enabling direct and rapid linkage between the ADAPA scoring engine and the data repository (whether MySQL, Microsoft SQL Server, Oracle, PostgreSQL or many other databases).

Data geeks rejoice! (and keep reading below for more details) Those readers who are not data geeks can rest assured that the data geeks have this problem solved. Zementis and Pentaho just made your lives easier.

Best Regards,Zementis-Logo hi-res

Michael Zeller
CEO, Zementis Inc.


Pentaho Data Integration and ADAPA: Multi-database Support That Maximizes Your Flexibility

Zementis customers using our ADAPA product frequently ask how they can integrate with a relational database. For our customers who use Pentaho for their data integration and analytics needs, the answer is simple: score your data in ADAPA using the Pentaho Data Integration (PDI) tool.

pentahoADAPA makes this simple, using its REST API and PDI to interface seamlessly with the data to be scored, whether it resides in MySQL, Microsoft SQL Server, Oracle or PostgreSQL. The setup is quick and simple, using an intuitive point-and-click interface to manage the entire workflow – retrieving the data, scoring it through ADAPA and then saving the results. The process is designed to be user-friendly and doesn’t require writing any code. In just a few hours, users can link their relational databases to a scoring engine and generate predictive results.

For more details, please refer to “Scoring data with ADAPA using Pentaho Data Integration” or to see it in action, please review one of the following videos that illustrate the simplicity of using ADAPA to score data against a neural network model with Pentaho’s PDI.


MySQL and ADAPA: Turbo-charging Predictive Analytics Scoring

MySQL users enjoy the scalability, performance and cost-efficiency benefits of the world’s most popular open source database. With Zementis, they also benefit from the ease and speed of scoring predictive models using a MySQL database as a client to the ADAPA scoring engine. ADAPA’s REST API creates a seamless interface between ADAPA and MySQL using an HTTP-based UDF, driving predictive models and making their predictions more accurate, relevant and timely.

The benefits of using ADAPA for predictive analytics scoring with MySQL include accelerated time-to-insight for predictive analytics and more efficient data science:

  • Less time required to deploy predictive modelsmysql-500x308
  • Little or no involvement of the IT organization
  • More time for data scientists to devote to model development and testing

In addition, it results in a lower operating cost structure for a predictive analytics capability:

  • Lower utilization of human capital
  • Higher utilization of existing IT assets

By simply using SQL and User Defined Functions (UDFs), users can easily execute complex predictive analytics models directly from one of the most commonly used database solutions, score the records, and write the results back into a database table. Zementis helps MySQL users extract the full value of data for their predictive analytics needs.

Zementis recognizes the power of leveraging an organization’s entire data set to drive predictive analytics, including data residing in relational databases such as MySQL. To see ADAPA in action with MySQL, including how to install MySQL-UDF-HTTP and write functions and triggers, please review the following resources:

Blog: Using MySQL as a Client to the ADAPA Scoring Engine
Video: Using MySQL UDF Trigger Update with ADAPA through REST API

Copyright © 2017 Zementis / Software AG. All Rights Reserved.