H2O

来自开放百科 - 灰狐
(版本间的差异)
跳转到: 导航, 搜索
(安装)
(图集)
 
(未显示1个用户的21个中间版本)
第9行: 第9行:
 
==简介==
 
==简介==
 
H2O - the killer app for Spark
 
H2O - the killer app for Spark
 +
 +
[[image:sparkling-killer-app.png]]
  
 
==版本==
 
==版本==
*[https://github.com/h2oai/h2o-3 h2o-3]
+
*[https://github.com/h2oai/h2o-3 h2o-3] [https://github.com/h2oai/h2o-3/tree/master/h2o-docs/src h2o-3文档] [http://docs.huihoo.com/h2o/3/ 在线文档]
*[https://github.com/h2oai/h2o-2 h2o-2]
+
*[https://github.com/h2oai/h2o-2 h2o-2] [https://github.com/h2oai/h2o-tutorials h2o-tutorials]
  
 
==安装==
 
==安装==
第21行: 第23行:
 
  http://localhost:54321
 
  http://localhost:54321
  
==指南==
+
==h2o.js==
 +
[https://github.com/h2oai/h2o.js h2o.js]: [[Node.js]] bindings to H2O
 +
 
 +
==H2O Flow==
 +
[https://github.com/h2oai/h2o-flow H2O Flow]
 +
 
 +
==机器学习==
 +
*[https://github.com/h2oai/deepwater/ Deep Water] [[Deep learning]] in H2O using Native [[GPU]] Backends 
 +
*[http://docs.huihoo.com/h2o/3/data-science/deep-learning.html H2O Deep Learning 文档]
 +
 
 +
==项目==
 +
[https://github.com/h2oai/awesome-h2o Awesome H2O] [[文件:awesome.png]]
 +
 
 +
==文档==
 +
*[http://docs.huihoo.com/knime/summits/knime-fall-summit-2017-austin/Integrating-high-performance-machine-learning-H2O-and-KNIME.pdf Integrating high performance machine learning: H2O and KNIME]
 +
*[http://docs.huihoo.com/h2o/h2o-meetups/ 更多幻灯片>>>]
  
 
==图集==
 
==图集==
 
<gallery>
 
<gallery>
 +
image:Gartner-2018-Magic-Quadrant-for-Data-Science-and-Machine-Learning.png|Gartner魔力象限
 +
image:why-h2o.png|Why H2O
 +
image:H2O-High-Level-Architecture.png|架构
 +
image:H2O-Distributed-Algorithms.png|分布式算法
 
image:H2O-Architecture-Slide.png|架构
 
image:H2O-Architecture-Slide.png|架构
 
image:h2o-software-stack.png|软件堆栈
 
image:h2o-software-stack.png|软件堆栈
 
image:h2o-sparkling-water-design.png|Sparkling Water
 
image:h2o-sparkling-water-design.png|Sparkling Water
 
image:h2o-sparkling-water-cluster.png|Sparkling Water集群
 
image:h2o-sparkling-water-cluster.png|Sparkling Water集群
 +
image:h2o-frame-data-structures.jpg|H2OFrame
 +
image:h2o-spark-workflow.jpg|工作流
 +
image:h2o-run-glm-from-r.png|GLM处理流
 +
image:h2o-flow.png|H2O Flow
 +
image:h2o-flow-model.png|模型
 +
image:h2o-steam.png|Steam
 +
image:h2o-storm.png|H2O on Storm
 
</gallery>
 
</gallery>
  
第42行: 第70行:
 
[[category:python]]
 
[[category:python]]
 
[[category:spark]]
 
[[category:spark]]
 +
[[category:storm]]
 
[[category:hadoop]]
 
[[category:hadoop]]

2018年3月17日 (六) 08:14的最后版本

Wikipedia-35x35.png 您可以在Wikipedia上了解到此条目的英文信息 H2O Thanks, Wikipedia.

h2o = fast statistical, machine learning & math runtime for big data

H2O是一个开源分布式的内存处理引擎用于机器学习和大数据,它拥有一个令人印象深刻的数组的算法,支持RPythonJava语言,同时它也可以作为Apache Spark在后端的执行引擎。

使用H2O的最佳方式是把它作为R环境的一个大内存扩展,R环境并不直接作用于大的数据集,而是通过扩展通讯协议例如REST API与H2O集群通讯,H2O来处理大量的数据工作。

目录

[编辑] 简介

H2O - the killer app for Spark

Sparkling-killer-app.png

[编辑] 版本

[编辑] 安装

curl -o h2o.zip http://download.h2o.ai/versions/h2o-3.8.1.4.zip
unzip h2o.zip
cd h2o-3.8.1.4
java -jar h2o.jar
http://localhost:54321

[编辑] h2o.js

h2o.js: Node.js bindings to H2O

[编辑] H2O Flow

H2O Flow

[编辑] 机器学习

[编辑] 项目

Awesome H2O Awesome.png

[编辑] 文档

[编辑] 图集

[编辑] 链接

分享您的观点
个人工具
名字空间

变换
操作
导航
工具箱