TAP Analytics Toolkit (ATK)

TAP is evolving from an integrated open source software platform to a set of open source software components under Apache 2.0 license.

TAP ATK (Analytics Toolkit) is one of the first core software components to be individually published to GitHub. TAP ATK enables data scientists and application developers to accelerate development of various analytics, from diagnostic to predictive to prescriptive, resulting in rapid application creation and faster business insights.

The software:

  • Facilitates data ingestion, preparation and analytics with distributed analytics and parallel processing
  • Improves analytics modeling process with optimized computation ML and DL algorithms, graph operations, scoring, and pipelines
  • Integrates with industry leading software frameworks such as Spark, Tensorflow and Superset to  expedite application development

TAP ATK leverages Apache Spark, Intel Data Analytics Acceleration Libraries (DAAL) and Intel Math Kernel Libraries (MKL) for optimized distributed analytics and parallel processing on Intel processors.

To accelerate the analytics modeling process, TAP ATK includes Intel-optimized Machine Learning, Deep Learning and Graph Algorithms and operations, Hadoop and Spark optimizations and advanced scoring engine with pipelines.

TAP ATK APIs enable DL and Visualization techniques integrated with a multiplicity of software frameworks such as Tensorflow and Superset.

For more information on TAP ATK