libSkylark documentation
Sketching-based Matrix Computations for Machine Learning
Contents
::
Introduction
»
Sketching-based Matrix Computations for Machine Learning
¶
Contents:
Introduction
Goals
Sketching and Sampling Methods
License and Copyright
Quick Start Guide
Installing libskylark from conda channels
Installing libskylark from a single-file installer
Vagrant
Command-line Usage
Cluster of vagrant-controlled VMs
Running Vagrant on AWS
Running MPI on Hadoop/Yarn
Examples of Library Usage
Building from Source
Getting the Source Code
libSkylark Software Dependencies
Building libSkylark
Using libSkylark in Your Project
Software Pitfalls
Complete BGQ Installation Instructions
Base Layer
Random Values
libSkylark Context
Local sparse matrices
Random Dense Matrices
Cross matrix-type GEMM and other linear algebra routines
Sketching Layer
Introduction
Overview of High-performance Distributed Sketching Implementation
libSkylark’s Sketching Layer
Numerical Linear Algebra Primitives
Randomized Linear Least-Squares
Randomized Singular Value Decomposition
Condition Number Estimation
Machine Learning
Randomized Kernel Methods
Graph Computations
IO in libSkylark
IO from C++
IO from Python
Sketch Serialization
Contributing
License Agreement
Coding style
Indices and tables
¶
Index
Module Index
Search Page
Contents
::
Introduction
»