Parallel Iterative Algorithm (SGD) on Hadoop's YARN framework
- Built on top of BSP-style computation framework "Iterative Reduce" (Hadoop / YARN)
- Uses Mahout's implementation of Stochastic Gradient Descent (SGD) as basis for worker process
Slides From Hadoop World 2012:
Knitting Boar is an experimental machine learning application which parallelizes Mahout's Stochastic Gradient Descent on top of a new YARN based framework for Hadoop called Iterative Reduce
- Intro to Knitting Boar
- [Quick Start] (https://github.com/jpatanooga/KnittingBoar/wiki/Quick-Start)
- [Frequently Asked Questions] (https://github.com/jpatanooga/KnittingBoar/wiki/FAQ)
- [Command Line Usage] (https://github.com/jpatanooga/KnittingBoar/wiki/Command-Line-Usage)
- [Knitting Boar Internals] (https://github.com/jpatanooga/KnittingBoar/wiki/Knitting-Boar-Internals)
- [Iterative Reduce] (https://github.com/jpatanooga/KnittingBoar/wiki/Iterative-Reduce)
- [Outstanding Issues] (https://github.com/jpatanooga/KnittingBoar/wiki/Outstanding-Issues)
- [Parallel SGD Resources] (https://github.com/jpatanooga/KnittingBoar/wiki/Parallel-SGD-Resources)