- Product
- Product Resource
Vertica Announces Vertica 12 for Future-Proof Analytics
Latest version of analytics database enables more deployment flexibility, advanced analytics, and enhanced machine learning
-
- Partners
- Partners Resource
Vertica Inside – Embedded Analytics at Scale
Seize the huge growth opportunity for OEM software developers
-
- Resources
- Resources Menu Resource
GigaOm Data Warehouse Report
Vertica receives multiple Exceptional ratings for key criteria, market and user categories and recognition for future-looking product development
-
Machine learning is gaining popularity as an essential way of not only identifying patterns and relationships, but also predicting outcomes. This is creating a fundamental shift in the way businesses are operating—from being reactive to being proactive.
This shift can help reduce customer churn, predict mechanical equipment failures, detect fraud, or bring entirely new products to market. Unfortunately, the growing volume of data has increased the complexity of building predictive models, since few tools are capable of processing these massive datasets at the speed of business.
To address this problem, more organizations are turning to platforms that leverage in-database machine learning in order to reduces volume constraints of traditional tools and enable discovery of patterns buried in ever-larger datasets.
Is your organization evaluating in-database analytics platforms to enable new machine learning capabilities? Consider these 10 critical components:
1. Scale out to hundreds of TBs or PBs of data
- The platform must leverage a modern, massively parallel processing architecture capable of scaling from Terabytes to Petabytes of data storage and analysis.
- Enables the preparation, training, scoring and deploying of machine learning models at the scale of today’s data volumes.
2. Train and score machine learning models with extreme speed
- The platform must be able to build machine learning models on massive data sets in seconds or minutes rather than hours or days.
- Enables data science and analyst teams to quickly iterate on model development and deploy new predictive analytics projects in less time than traditional statistical analysis tools.
3. Integrate with a wide range of ecosystem solutions and services
- The platform must integrate with common ETL and BI/visualization tools through the formation of collaborative ecosystem partnerships and product certifications.
- Enables users to extend the value of an in-database analytics platform with access to familiar applications across the entire data and analytics stack.
4. Extend the reach of machine learning to non-data scientists
- The platform must empower data analysts and other stakeholders to build and operationalize predictive analytics projects using simple SQL calls.
- Enables companies to expand the pool of resources available for predictive analytics projects by putting machine learning in the hands of more employees.
5. Include enterprise features out of the box
- The platform must support common enterprise-grade software features, such as robust security, ACID compliance, high availability, professional support, and more.
- Provides assurance that the platform can be a trusted, integral component of the company’s IT operations and product delivery.
Vertica In-database Machine Learning
Vertica’s in-database machine learning supports the entire predictive analytics process with massively parallel processing and a familiar SQL interface, allowing data scientists and analysts to embrace the power of Big Data and accelerate business outcomes with no limits and no compromises.
6. Support data preparation, evaluation and deployment
- The platform must provide features that support the entire machine learning process with functions for data preparation, model creation, evaluation and deployment.
- Reduces the complexity of end-to-end machine learning management with a simple SQL-based advanced analytics platform.
7. Available on-premises and in the clouds
- The platform must provide flexibility to deploy anywhere – on commodity hardware, across multiple clouds (AWS, Azure, Google) and natively on any Hadoop distribution.
- Ensures freedom from underlying infrastructure so organizations can avoid hardware vendor lock-i and analyze data with the highest performance, regardless of where that data resides.
8. Integrate with the leading open source tools
- The platform must be interoperable with the most widely adopted open source solutions in data analytics and machine learning, such as Apache Spark, Kafka and Hadoop.
- Protects your open source infrastructure investments and ensures a well integrated stack of technology solutions for machine learning and predictive analytics.
9. Bring machine learning and predictive analytics to the data
- The platform must provide a means of storing massive volumes of data so machine learning functions can be run in-database, where that data resides.
- Accelerates time-to-value on predictive analytics projects by eliminating the need to move data across tools, or compromise model accuracy with down sampling.
10. Support familiar machine learning programming languages
- The platform must allow development of user-defined extensions (UDx) in popular machine learning programming languages, such as R and Python.
- Increases the power and flexibility of procedural code by executing algorithms in parallel, across nodes, on large datasets.