SAN JOSE, CA--(Marketwired - Feb 1, 2017) - Skytree, the leader in Enterprise Machine Learning on big data, announces the release of Skytree 16.0 and Skytree's Machine Learning as a Service offering. We continue our trend of increasing ease of use via unprecedented automation, further enabling non-data scientist users to access the power of enterprise grade machine learning, gain insights, and to add value to their business.
Skytree Smart Search continues to deliver industry-leading automation for finding the most accurate models. This is accomplished in much fewer iterations than brute-force grid searches, by performing a mathematically optimal exploration of the hyper-parameter space. Skytree Smart Search delivers highly tuned models with comprehensive cross-validation in every step, and delivers reliable execution with automatic resource estimation for the powerful Skytree distributed algorithm implementations that are highly optimized for scale, performance and accuracy. Skytree AutoModel provides all of the benefits of Skytree Smart Search across a unified search space of machine learning methods and their hyper-parameters. Skytree automation using Skytree Smart Search and Skytree AutoModel allow data scientists and data analysts to find better models, in significantly less time, with much less user effort when compared to brute force or manually configured grid searches, manual selection of methods, and manual tuning of model parameters.
More on Skytree Smart Search and Skytree AutoModel can be found at http://www.skytree.net/2015/05/29/automatic-method-and-parameter-selection
The two major changes in Skytree 16.0 are: 1) the addition of automatic featurization, and 2) a major redesign of the Skytree GUI.
The first step in enabling feature engineering comes from the added functionality of transform snippets to the Skytree Platform, and the second step in delivering automatic featurization comes from automated heuristic analysis of datasets and determination of the appropriate set of transform snippets to be applied to them to improve model accuracy. Snippets allow users to perform the arbitrary data preparation demanded by real world data flows while taking advantage of the scalability of Spark without the requirement to know the Spark language. Furthermore, with snippets added to the system, GUI users can use them without needing to know any coding at all.
The graphical interface in Skytree's Machine Learning platform has been overhauled for Skytree 16.0 to give a more fluid and user friendly experience, both in the way information is displayed, and for navigation between different parts of a project. Additional tools have been added to help the user to get to their optimal model faster and with greater ease.
More on what's new in Skytree 16.0 can be found in the following:
Blog Post: http://www.skytree.net/2017/01/18/whats-new-in-skytree-16-0/#more-5521
With the new release of Skytree 16.0, Skytree is now able to easily offer Machine Learning as a Service. Skytree's Machine Learning as a Service delivers a fully tuned Turnkey Machine Learning Model on your data for as little as $25K for the initial model.
In the Machine Learning as a Service offering, Skytree provides everything: the Compute Platform (via AWS), the Data Scientist staff, and the Skytree Machine Learning enterprise platform leveraging the Skytree patented Machine Learning automation. All that is needed is Machine Learning ready data and the business case.
Implementing Machine Learning often requires a lot of time, complex technology, and specialized expertise. Only after you hire a Data Science team, implement the proper technology, and spend months experimenting with different algorithms and models will you know if Machine Learning has given you the results you need.
Machine Learning is an iterative process. Oftentimes, the insights from the initial training of the Machine Learning Model will contribute a significant positive impact to your business immediately. The continued tuning, iteration and improvement of the Machine Learning Model then continuously improves the accuracy and insights over time.
"At Skytree, we have created a way to produce the initial Machine Learning Model with minimum time and resource commitment by the client by leveraging Skytree's unique automation. We want people to experience that initial impact in the quickest way possible," says Dr. Alexander Gray, CEO & Co-Founder of Skytree.
This service leverages the Skytree Data Science team and our patented automation from the Skytree Machine Learning Platform hosted on Amazon Web Services to quickly produce a fully tuned Machine Learning Predictive Model and all of the related predictive insights.
Once a client's model is created, the client will have access to a personal portal where they can interact with their data and their optimized Machine Learning Model and insights. Clients will be able to immediately interpret these insights and instantly apply them to their business. The model will also immediately be available to export and use in production environment.
Contact us for more information at: http://pages.skytree.net/Machine-Learning-as-a-Service.html