Skytree®, a leader in enterprise-grade machine learning on big data, announced the release of Skytree 15.2. This release enhances Skytree’s high performance, scalable machine-learning platform with streamlined data preparation for unstructured text data. The latest release also improves model interpretability, heightens security and extends support for fusing structured and unstructured data.
Building machine-learning models with both unstructured and structured data fundamentally delivers more accurate models. Until now, it has been difficult, if not impossible to build machine-learning models from large-scale diverse data sets, despite the proliferation of unstructured data, like text data, within enterprises. Skytree 15.2 is the first tool able to break up the schema of text and extract meaningful tokens for analysis, automatically, and in a highly secure and available environment,” said Alexander Gray, Ph.D., chief technology officer and co-founder of Skytree. “The latest release of Skytree makes it easy to apply advanced machine learning to a wealth of unstructured data.”
Unstructured text data makes up nearly 80 percent of the typical enterprise’s data but presents significant challenges for data scientists, often requiring an inordinate amount of time to structure and prepare data for analysis. Skytree 15.2 simplifies the featurization of text data from sources that include Word, Excel, PowerPoint, PDF, raw text, tweets and Web pages. Machine-learning models can be built with standalone or fused structured and unstructured data. Skytree also delivers an even more secure LDAP authentication capability with SSL for single sign-on.
Machine learning is the next frontier for big data, enabling analytical insights on vast and varied data sources. Skytree is making smart moves to extend its platform both in terms of reach and maturity, supporting more ready access to data and more governance and availability to meet enterprise requirements,” says Nik Rouda, senior analyst of Enterprise Strategy Group.
Skytree 15.2 enhances the ability to administer and deploy in a customer’s environment with an easy to use administrative console, allowing organizations to easily manage users, as well as better manage and allocate big data resources. The latest release also enables a high availability solution with stateful failover capabilities on Hadoop edge nodes. With Skytree, IT organizations can meet tight Service Level Agreements (SLAs), and data scientists can reduce the time to deliver models when hardware nodes fail.
Skytree is actively working with customers in key industries, including financial services, insurance, telecommunications, retail, technology and government to provide support in business-critical functions such as fraud detection, churn prevention, risk analysis, predictive maintenance, network analytics, targeting and recommendations.
Sign up for the free insideBIGDATA newsletter.
Speak Your Mind