News
AWS SageMaker Algorithms and Frameworks Get Refresh
Amazon Web Services (AWS) is rolling out several improvements to its SageMaker machine learning modeling platform, the company announced last week.
The improvements apply to the program's built-in BlazingText, DeepAR and Linear Learning algorithms.
For example, AWS said it improved the accuracy and ease of using DeepAR's algorithms for forecasting so that "missing values are now handled within the model." Its DeepAR algorithm also now supports seasonality patterns and other "custom time-varying features," as well as multiple groupings of time series.
BlazingText has been improved, according to the company, with the addition of a Word2Vec algorithm designed to optimize usage of GPU hardware.
Other improvements include "meaningful vectors for out-of-vocabulary (OOV) words that do not appear in the training dataset" and new support for "high-speed multi-class and multi-label text classification," among other improvements.
Users of Linear Learning will now have multi-class classification and other improved ways of sorting and classifying data.
SageMaker also now supports Chainer 4.1, offering pre-configured containers within which users will find Layer-wise Adaptive Rate Scaling (LARS), which AWS says improves the training of networks with "large batch sizes."
The above changes are currently rolling out across AWS instances worldwide.
About the Author
Becky Nagel is the vice president of Web & Digital Strategy for 1105's Converge360 Group, where she oversees the front-end Web team and deals with all aspects of digital projects at the company, including launching and running the group's popular virtual summit and Coffee talk series . She an experienced tech journalist (20 years), and before her current position, was the editorial director of the group's sites. A few years ago she gave a talk at a leading technical publishers conference about how changes in Web browser technology would impact online advertising for publishers. Follow her on twitter @beckynagel.