Apache Mahout (http://mahout.apache.org) is a distributed linear algebra framework that includes a mathematically expressive domain-specific language (DSL). It is designed to aid mathematicians, statisticians, and data scientists to quickly implement numerical algorithms while focusing on the mathematical concepts in their work, rather than on code syntax. Mahout uses an extensible plug-in interface to systems such as Apache Spark and Apache Flink.
Mahout was founded as a sub-project of Apache Lucene in late 2007 and was promoted to a top-level Apache Software Foundation (ASF) (ASF 2017) project in 2010 (Khudairi 2010). The goal of the project from the outset has been to provide a machine learning framework that was both accessible to practitioners and able to perform sophisticated numerical computation on large data sets.
Mahout has undergone two major stages of architecture design. The first versions relied on the Apache Hadoop MapReduce framework, a...
KeywordsApache Mahout Compatible GPUs Apache Spark HTTP Access Logs Hadoop
- ASF (2017) Welcome to the apache software foundation! https://www.apache.org
- Khudairi S (2010) The apache software foundation blog. https://blogs.apache.org/foundation/entry/the_apache_software_foundation_announces4
- PMC AM (2015) Apache mahout 0.10.0 release notes. http://mahout.apache.org/release-notes/Apache-Mahout-0.10.0-Release-Notes.pdf
- PMC AM (2017) Apache mahout 0.13.0 release notes. https://mail-archives.apache.org/mod_mbox/www-announce/201704.mbox/%3CCANg8BGBe+WwdZC6z6BAm3hqTOMjA2ma76y0dig0Jf5LHtgF56g@mail.gmail.com%3E