View source for Software:H2O

{{Infobox software
| title                  = H2O
| name                   = H2O (software)
| logo                 =H2O_logo_from_H2O.ai.png <!-- FAIR USE. Not for use elsewhere -->
| logo size              =
| logo alt               = The corporate logo of H2O
| logo caption           =
| screenshot             = <!-- Image name is enough -->
| screenshot size        =
| screenshot alt         =
| caption                =
| collapsible            =
| author                 = H2O.ai
| developer              = H2O.ai
| released               = {{Start date and age|2011}}
| discontinued           =
| latest release version = 3.20.0.8
| latest release date    = {{Start date and age|2018|08|01}}
| status                 = Active
| programming language   = H2O (written in [[Java (programming language)|Java]], [[Python (programming language)|Python]], and [[R (programming language)|R]])<ref name="Harris" >{{harvtxt|Harris|2012}}</ref><ref name="Novet">{{harvtxt|Novet|2014}}</ref><ref name="Rec">{{cite web|publisher=H2O.ai|website=0xdata.com|url=http://0xdata.com/product/recommended-systems-for-h2o/|title=Recommended systems for H2O|ref=harv|date=May 2015|access-date=2015-06-01|archive-url=https://web.archive.org/web/20150530010057/http://0xdata.com/product/recommended-systems-for-h2o/|archive-date=2015-05-30|dead-url=yes|df=}}</ref>
| operating system       = [[Software:Linux|Linux]], [[Software:MacOS|macOS]], and [[Software:Microsoft Windows|Microsoft Windows]]
| platform               = [[Software:Apache Hadoop|Apache Hadoop]] [[Software:Apache Hadoop#Hadoop distributed file system|Distributed File System]]; Amazon EC2, [[Software:Google Compute Engine|Google Compute Engine]], and [[Software:Microsoft Azure|Microsoft Azure]].
| size                   =
| language               = English
| language count         = <!-- Number only -->
| language footnote      =
| genre                  = [[Big data|big]] data analytics, [[Machine learning|machine learning]], [[Statistical learning theory|statistical learning theory]]<ref name="Hardy" >{{harvtxt|Hardy|2014}}</ref>
| license                = Apache license 2.0<ref>https://github.com/h2oai/h2o-2/blob/master/LICENSE.txt</ref>
| website                = {{URL|http://www.h2o.ai/}}
| standard               = [[Company:Databricks|Databricks]] certified on [[Software:Apache Spark|Spark]].<ref name="Rec"/>
| AsOf                   = 1 June 2015
}}

'''H2O''' is [[Open-source software|open-source software]] for [[Big data|big-data]] [[Data analysis|analysis]]. It is produced by the company H2O.ai. H2O allows users to fit thousands of potential models as part of discovering patterns in data.

The H2O software runs can be called from the [[R (programming language)|statistical package R]], [[Python (programming language)|Python]], and other environments. It is used for exploring and analyzing datasets held in [[Cloud computing|cloud computing]] systems and in the [[Software:Apache Hadoop|Apache]] Hadoop Distributed File System as well as in the conventional operating-systems [[Software:Linux|Linux]], [[Software:MacOS|macOS]], and [[Software:Microsoft Windows|Microsoft Windows]]. The H2O software is written in [[Java (programming language)|Java]], [[Python (programming language)|Python]], and [[R (programming language)|R]]. Its graphical-user interface is compatible with four browsers: Chrome, Safari, Firefox, and [[Software:Internet Explorer|Internet Explorer]].

==H2O==

The H2O project aims to develop an analytical interface for cloud computing, providing users with tools for data analysis.<ref name="Harris" /> The software is open-source and freely distributed. The company receives fees for providing customer service and customized extensions.

===Mining of big data===
{{See also|Data mining|Machine learning}}{{machine learning bar}}[[Big data]]sets are too large to be analyzed using traditional software like [[R (programming language)|R]]. The H2O software provides data structures and methods suitable for big data. H2O allow users to analyze and visualize whole sets of data without using the Procrustean strategy of studying only a small subset with a conventional statistical package.<ref name="Novet" /> H2O's statistical algorithms includes [[K-means clustering]], [[Generalized linear model|generalized linear model]]s, [[Random forest|distributed random forests]], [[Gradient boosting|gradient boosting machines]], [[Naive Bayes classifier|naive bayes]], [[Principal component analysis|principal component analysis]], and [[Low-rank approximation|generalized low rank models]].<ref name="CRAN">{{citation|title =h2o: R Interface for H2O|author =Aiello, Spencer|author2 =Tom Kraljevic|author3 =Petr Maj|others=with contributions from the 0xdata team|year=2015|number=3.0.0.12
|url=https://cran.r-project.org/web/packages/h2o/index.html
|ref=harv|series=Contributed Packages|publisher=The R Project for Statistical Computing|website=The Comprehensive R Archive Network (CRAN)}}</ref>

H2O is also able to run on Spark.<ref>{{Cite web|url=http://docs.h2o.ai/h2o/latest-stable/h2o-docs/faq.html#sparkling-water|title=FAQ — H2O 3.10.2.1 documentation|website=docs.h2o.ai|language=en|access-date=2017-01-28}}</ref>

====Iterative methods for real-time problems====

H2O uses [[Iterative method|iterative method]]s that provide quick answers using all of the client's data. When a client cannot wait for an optimal solution, the client can interrupt the computations and use an approximate solution.<ref name="Harris" /> In its approach to [[Deep learning|deep learning]],<ref name="Novet" /><ref name="CRAN" /><ref>"Prediction of IncRNA using Deep Learning Approach". Tripathi, Rashmi; Kumari, Vandana; Patel, Sunil; Singh, Yashbir; Varadwaj, Pritish. ''International Conference on Advances in Biotechnology (BioTech)''. Proceedings: 138-142. Singapore: Global Science and Technology Forum. (2015)</ref> H2O divides all the data into subsets and then analyzing each subset simultaneously using the same method. These processes are combined to estimate parameters by using the Hogwild scheme,<ref>[http://h2o-release.s3.amazonaws.com/h2o/rel-shannon/12/docs-website/h2o-docs/index.html#Data%20Science%20Algorithms-GLM-GLM%20Algorithm Description of the iterative method for computing maximum-likelihood estimates for a generalized linear model].</ref> a parallel [[Stochastic gradient descent|stochastic gradient]] method.<ref>{{cite journal|title=Hogwild: A Lock-Free Approach to Parallelizing Stochastic Gradient Descent|author =Benjamin Recht|author2 =Re, Christopher|author3 =Wright, Stephen|author4 =Feng Niu|last-author-amp =yes
|journal=Advances in Neural Information Processing Systems|volume=24|editor=J. Shawe-Taylor|editor2=R.S. Zemel|editor3=P.L. Bartlett|editor4=F. Pereira|editor5=K.Q. Weinberger|pages=693–701|year=2011|publisher=Curran Associates, Inc.
|url=http://papers.nips.cc/paper/4390-hogwild-a-lock-free-approach-to-parallelizing-stochastic-gradient-descent.pdf
|ref=harv}} [https://www.eecs.berkeley.edu/~brecht/papers/hogwildTR.pdf Recht's PDF]</ref> These methods allow H2O to provide answers that use all the client's data, rather than throwing away most of it and analyzing a subset with conventional software.

===Software===
<!-- Following the guidelines of the article's WikiProjects and the software information-box template, this section provides the information requested by the template. Citations can be expanded to include facts from Novet and from Harris, but this would just clutter the article. It is simplest and cleanest to use H2O.ai's recommendations as the uniform source. -->

====Programming languages====

The H2O software has an interface to the following programming languages: [[Java (programming language)|Java]] (6 or later), [[Python (programming language)|Python]] (2.7.x, 3.5.x), [[R (programming language)|R]] (3.0.0 or later) and [[Scala (programming language)|Scala]] (1.4-1.6).<ref name="Novet" /><ref name="Rec" />

====Operating systems====
The H2O software can be run on conventional operating-systems: [[Software:Microsoft Windows|Microsoft Windows]] (7 or later), Mac [[Software:OS X|OS X]] ([[Software:OS X Mavericks|10.9]] or later), and [[Software:Linux|Linux]] (Ubuntu 12.04 <!-- not "or later" 1 June 2015 -->; RHEL/[[Software:CentOS|CentOS]] 6 or later),<ref name="Rec"/> It also runs on big-data systems, particularly [[Software:Apache Hadoop|Apache Hadoop]] [[Software:Apache Hadoop#Hadoop distributed file system|Distributed File System]] (HDFS), several popular versions: [[Company:Cloudera|Cloudera]] (5.1 or later), [[Company:MapR|MapR]] (3.0 or later), and [[Company:Hortonworks|Hortonworks]] (HDP 2.1 or later). It also operates on [[Cloud computing|cloud computing]] environments, for example using Amazon EC2, [[Software:Google Compute Engine|Google Compute Engine]], and [[Software:Microsoft Azure|Microsoft Azure]]. The H2O Sparkling Water software is [[Company:Databricks|Databricks]]-certified on [[Software:Apache Spark|Apache Spark]].<ref name="Rec"/>

====Graphical user interface and browsers====

Its graphical user interface is compatible with four browsers (unless specified, in their latest versions {{as of|2015|June|1|lc=y}}): Chrome, Safari, Firefox, [[Software:Internet Explorer|Internet Explorer]] ([[Software:Internet Explorer 10|IE10]]).<ref name="Rec"/>

==Notes==
{{Reflist}}

==References==
* {{cite news|last=Gage|first=Deborah|title=Platfora founder goes in search of big-data answers|url=https://blogs.wsj.com/venturecapital/2013/04/15/platfora-founder-goes-in-search-of-big-data-answers/|date=15 April 2013|newspaper=Wall Street Journal|ref=harv|accessdate=2 June 2015}}
* {{citation
|url=http://fortune.com/2014/08/03/meet-fortunes-2014-big-data-all-stars/
|journal=Fortune|first1=Robert|last1=Hackett
|title=Arno Candel, physicist and hacker, 0xdata
|editor-first2=Robert|editor-last2=Hackett|series=Meet Fortune's 2014 Big Data All-Stars|ref=harv|accessdate=2 June 2015|editor-last1=Nusca|editor-first1=Andrew|editor-last3=Gupta|editor-first3=Shalene|date=3 August 2014}}
* {{cite news|title=Valuable humans in our digital future|first=Quentin|last=Hardy|date=3 May 2014|newspaper=New York Times
|url=http://bits.blogs.nytimes.com/2014/05/03/valuable-humans-in-our-digital-future/?
|ref=harv|accessdate=1 June 2015}}
* {{cite journal|title=How 0xdata wants to help everyone become data scientists|first=Derrick|last=Harris|date=14 August 2012|ref=harv
|url=https://gigaom.com/2012/08/14/how-0xdata-wants-to-help-everyone-become-data-scientists/|journal=Gigaom Research|accessdate=1 June 2015}}
* {{cite journal|title=0xdata takes $8.9M and becomes H2O to match its open-source machine-learning project|date=7 November 2014|first=Jordan|last=Novet|journal=VentureBeat
|url=https://venturebeat.com/2014/11/07/h2o-funding/|ref=harv|accessdate=1 June 2015}}

==External links==
* {{URL|https://www.h2o.ai/h2o/|H2O software}} at {{URL|http://0xdata.com/ |H2O.ai (formerly 0xdata)}}
* {{URL|https://github.com/0xdata/h2o|Github repository of H2O software}}
* {{URL|https://www.youtube.com/channel/UCk6ONJlPzjw3DohAeMSgsng|YouTube channel of H2O.ai}}

{{Statistical software}}
{{Data warehouse}}

[[Category:Free data analysis software]]
[[Category:Data mining and machine learning software]]
[[Category:Open-source cloud applications]]
[[Category:R (programming language)| ]]
[[Category:Cross-platform free software]]
[[Category:Data-centric programming languages]]
[[Category:Free statistical software]]
[[Category:Free data visualization software]]
[[Category:Hadoop]]

{{Sourceattribution|H2O (software)|1}}