http://venturebeat.com/2015/01/16/facebook-opens-up-about-more-of-its-cutting-edge-deep-learning-tools/
In the spirit of helping everyone do artificial intelligence more
efficiently, Facebook is giving away some of its seriously powerful deep
learning tools for free in the Torch open-source library.
More specifically, the social networking company is open-sourcing code that helps it run complex algorithms for deep learning, an increasingly popular type of artificial intelligence. The company will announce the news in a blog post today.
Deep learning involves training systems, called artificial neural
networks, on lots of information derived from audio, images, and other
inputs, and then presenting the systems with new information and
receiving inferences about it in response. Torch, an open-source library
that’s been around since 2002, contains a framework for building and
training neural networks.
Facebook is making several modules available via Torch, including one
with a convolutional neural network layer featuring highly customized
kernels — templates that slide over images to recognize certain objects —
that could help researchers and engineers at other companies catch up
with some of the performance improvements Facebook has made internally.
“For everyone out there, these kernels are faster than anything else
in the open-source community,” Soumith Chintala, a Facebook artificial
intelligence researcher and software engineer, told VentureBeat in an
interview.
What’s more, Facebook is releasing “containers” to help distribute
the work of training neural network across multiple graphics processors.
“You’ll get a good speed-up,” Chintala explained. And the supercharged
convolution layer code is as much as 23.5 times faster than the fastest
system that was publicly available until this point, Chintala wrote in
his blog post.
Facebook has previously contributed a wide variety of other software
to the rest of the world under open-source licenses. But these new
components demonstrate a serious commitment from within the Facebook
Artificial Intelligence Research arm the company first formed in 2013, with the arrival of deep learning luminary Yann LeCun. More recently, that group brought on Vladimir Vapnik, known for his work on the Support Vector Machine algorithm.
Google, Twitter, Spotify, Netflix, and others have been quickly
bringing in their own talent in the domain of deep learning. But
open-source contributions in the area have been uncommon, making
Facebook’s move notable.
Facebook could also get people outside of the company to improve the
Torch modules and thus find new researchers to bring aboard.
Hiring motivations aside, Chintala believes the Torch project holds
serious merit and that the new components should make it still more
powerful.
“It’s like building some kind of electronic contraption or, like, a
Lego set,” Chintala said. “You just can plug in and plug out all these
blocks that have different dynamics and that have complex algorithms
within them.”
At the same time, he said, Torch is actually not extremely difficult to learn — unlike, say, the Theano library.
“We’ve made it incredibly easy to use,” Chintala said. “We introduce
someone to Torch, and they start churning out research really fast.”
Friday, January 16, 2015
Thursday, January 15, 2015
Analisa Karya Seni Dengan Artificial Intelligence
The task of classifying pieces of fine art is hugely complex. When
examining a painting, an art expert can usually determine its style, its
genre, the artist and the period to which it belongs. Art historians
often go further by looking for the influences and connections between
artists, a task that is even trickier.
https://medium.com/the-physics-arxiv-blog/when-a-machine-learning-algorithm-studied-fine-art-paintings-it-saw-things-art-historians-had-never-b8e4e7bf7d3e
https://medium.com/the-physics-arxiv-blog/when-a-machine-learning-algorithm-studied-fine-art-paintings-it-saw-things-art-historians-had-never-b8e4e7bf7d3e
Tutorial pattern classification
A collection of tutorials and examples for solving and understanding machine learning and pattern classification tasks
https://github.com/rasbt/pattern_classification
https://github.com/rasbt/pattern_classification
Tuesday, January 13, 2015
Tawaran Kuliah Machine Learning Online di Coursera
Berikut ini beberapa tawaran kuliah online di bidang machine learning di Coursera. Belajarnya gratis, bayar kalau perlu sertifikatnya saja.
Artificial Intelligence Planning : https://www.coursera.org/course/aiplan
The course aims to provide a foundation in artificial intelligence techniques for planning, with an overview of the wide spectrum of different problems and approaches, including their underlying theory and their applications.
The Data Scientist’s Toolbox: https://www.coursera.org/course/datascitoolbox
Part of the Data Science Specialization »
Get an overview of the data, questions, and tools that data analysts and data scientists work with. This is the first course in the Johns Hopkins Data Science Specialization.
Practical Machine Learning: https://www.coursera.org/course/predmachlearn
Part of the Data Science Specialization »
Learn the basic components of building and applying prediction functions with an emphasis on practical applications. This is the eighth course in the Johns Hopkins Data Science Specialization.
Artificial Intelligence Planning : https://www.coursera.org/course/aiplan
The course aims to provide a foundation in artificial intelligence techniques for planning, with an overview of the wide spectrum of different problems and approaches, including their underlying theory and their applications.
The Data Scientist’s Toolbox: https://www.coursera.org/course/datascitoolbox
Part of the Data Science Specialization »
Get an overview of the data, questions, and tools that data analysts and data scientists work with. This is the first course in the Johns Hopkins Data Science Specialization.
Practical Machine Learning: https://www.coursera.org/course/predmachlearn
Part of the Data Science Specialization »
Learn the basic components of building and applying prediction functions with an emphasis on practical applications. This is the eighth course in the Johns Hopkins Data Science Specialization.
Wednesday, January 7, 2015
Kaggle InClass: Stanford's "Getting a Handel on Data Science" Winners' Report
http://blog.kaggle.com/2015/01/05/kaggle-inclass-stanfords-getting-a-handel-on-data-science-winners-report/
Excerpt:
On December 3rd, Stanford Data Mining & Analysis course (STATS202) wrapped up a heated Kaggle InClass competition, "Getting a Handel on Data Science". Beating out 92 other teams, "TowerProperty" came in first place. Below, TowerProperty outlines the competition and their journey to the top of the leaderboard. Kaggle InClass is provided free of charge to academics as a statistical and data mining learning tool for students. Instructors from any course dealing with data analysis may get involved!
Excerpt:
On December 3rd, Stanford Data Mining & Analysis course (STATS202) wrapped up a heated Kaggle InClass competition, "Getting a Handel on Data Science". Beating out 92 other teams, "TowerProperty" came in first place. Below, TowerProperty outlines the competition and their journey to the top of the leaderboard. Kaggle InClass is provided free of charge to academics as a statistical and data mining learning tool for students. Instructors from any course dealing with data analysis may get involved!
Monday, January 5, 2015
CIFAR-10 Competition Winners: Interviews with Dr. Ben Graham, Phil Culliton, & Zygmunt Zając
http://blog.kaggle.com/2015/01/02/cifar-10-competition-winners-interviews-with-dr-ben-graham-phil-culliton-zygmunt-zajac/
Dr. Ben Graham
Dr. Ben Graham is an Assistant Professor in Statistics and Complexity at the University of Warwick. With a categorization accuracy of 0.95530 he ranked first place.
Congratulations on winning the CIFAR-10 competition! How do you feel about your victory?
Thank you! I am very pleased to have won, and quite frankly pretty amazed at just how competitive the competition was.
When I first saw the competition, I did not think
the test error would go below about 8%. I assumed 32x32 pixels just
wasn't enough information to identify objects very reliably. As it
turned out, everyone in the top 10 got below 7%, which is roughly on a par with human performance.
Friday, January 2, 2015
Rattle: A Graphical User Interface for Data Mining using R
http://rattle.togaware.com/
See OnePageR for a suite of guides
for the Data Scientist using R.
Version 3.4.1 dated 2014-12-29.
> install.packages("rattle", repos="http://rattle.togaware.com", type="source")
$ wget http://togaware.com/access/rattle_3.4.1.tar.gz
Rattle (the R Analytical Tool To Learn Easily) presents statistical and visual summaries of data, transforms data into forms that can be readily modelled, builds both unsupervised and supervised models from the data, presents the performance of models graphically, and scores new datasets.
Errata
Brochure
"The book is very enjoyable reading and is filled with useful information. It is aimed at both students learning data mining and data miners who are using or learning R. People are likely to read it through the first time as a text book and then later use it as a reference, especially about the details of the R language. One of the strongest aspects of this book is Dr. Williams’ ability to simplify complex topics and explain them clearly. His descriptions of bagging and boosting are the most clear that I have ever read."
Bob Muenchen, author of R for SAS and SPSS Users, 30 June 2011
From Amazon:
For anyone looking to learn more about R, this would be a great introduction. Brian Tvenstrup (5 reviewers made a similar statement).
This book covers both Rattle, the R code that Rattle creates, and writing some R code from scratch. Robert A. Muenchen (2 reviewers made a similar statement).
In summary, I found the book very readable, the examples easy to follow, and the explanations and reasons for why different processes are done. G3N1U5 (2 reviewers made a similar statement).
Rattle runs under GNU/Linux, Macintosh OS/X, and MS/Windows. The aim is to provide an intuitive interface that takes you through the basic steps of data mining, as well as illustrating the R code that is used to achieve this. Whilst the tool itself may be sufficient for all of a user's needs, it also provides a stepping stone to more sophisticated processing and modelling in R itself, for sophisticated and unconstrained data mining.
Rattle is in daily use by Australia's largest team of data miners and by a variety of government and commercial enterprises, world wide. A number of international consultants also use Rattle in their daily business. Users include the Australian Taxation Office, Australian Department of Immigration, Ulster Bank, Toyota Australia, US Geological Survey, Carat Media Network, Institute of Infection and Immunity of the University Hospital of Wales, US National Institutes of Health, AIMIA Loyalty Marketing, Added Value, and many more. It is or has been used for teaching by the McMaster University, Australian National University, University of Canberra, University of Technology Sydney, Yale University, University of Southern Queensland, Revolution Analytics, Habin Institute of Technology Graduate School Shenzhen, and Many more.
The author of Rattle received a 2007 Australia Day Medallion, presented by the Commissioner of Taxation, for leadership and mentoring in Data Mining in the Australian Taxation Office and in Australia, and particularly cited the development and sharing of the Rattle system.
Rattle is also used to teach the practise of data mining. It was the primary tool of instruction for a Data Mining Workshop in Canberra, and at Harbin Institute of Technology, Shenzhen Graduate School (2006), and for the Australian National University's course on Data Mining (since 2006), University of Canberra (since 2010) and University of South Australia (since 2009). It has been used in courses at Yale University, University of Liège Belgium (since 2011), University of Wollongong (since 2010), University of Southern Queensland (since 2010), Australian Consortium for Social and Political Research (2011), University of Technology, Sydney (since 2012), Revolution Analytics (since 2012) and many others.
Further data mining resources are also available from Togaware.
Rattle is open source and freely available from Togaware. You can download Rattle and get familiar with its functionality without any obligation, except for the obligation to freely share! Organisations are also welcome to purchase Rattle, including support for installation and initial training, and ongoing data mining support. Email rattle@togaware.com for details.
Graham Williams (2011). Data Mining with Rattle and R: The Art of Excavating Data for Knowledge Discovery, Springer, Use R!.
or
Graham Williams (2009). Rattle: A Data Mining GUI for R, Graham J Williams, The R Journal, 1(2):45-55.
Rattle: A Graphical User Interface for Data Mining using R
See OnePageR for a suite of guides
for the Data Scientist using R.Version 3.4.1 dated 2014-12-29.
> install.packages("rattle", repos="http://rattle.togaware.com", type="source")
$ wget http://togaware.com/access/rattle_3.4.1.tar.gz
Rattle (the R Analytical Tool To Learn Easily) presents statistical and visual summaries of data, transforms data into forms that can be readily modelled, builds both unsupervised and supervised models from the data, presents the performance of models graphically, and scores new datasets.
Errata
Brochure
- Documentation
- Information
- Install
- Support
- Coding
- Other Resources
Reviews
"Rattle is a tab-oriented user interface that is similar to Microsoft Office’s ribbon interface. It makes getting started with data mining in R very easy. This book covers both Rattle, the R code that Rattle creates, and writing some R code from scratch. Therefore it will appeal to both people seeking the ease-of-use that is very much missing from R, and people looking to learn R programming.""The book is very enjoyable reading and is filled with useful information. It is aimed at both students learning data mining and data miners who are using or learning R. People are likely to read it through the first time as a text book and then later use it as a reference, especially about the details of the R language. One of the strongest aspects of this book is Dr. Williams’ ability to simplify complex topics and explain them clearly. His descriptions of bagging and boosting are the most clear that I have ever read."
Bob Muenchen, author of R for SAS and SPSS Users, 30 June 2011
From Amazon:
For anyone looking to learn more about R, this would be a great introduction. Brian Tvenstrup (5 reviewers made a similar statement).
This book covers both Rattle, the R code that Rattle creates, and writing some R code from scratch. Robert A. Muenchen (2 reviewers made a similar statement).
In summary, I found the book very readable, the examples easy to follow, and the explanations and reasons for why different processes are done. G3N1U5 (2 reviewers made a similar statement).
Information
Through a simple and logical graphical user interface based on Gnome, Rattle can be used by itself to deliver data mining projects. Rattle also provides an entry into sophisticated data mining using the open source and free statistical language R.Rattle runs under GNU/Linux, Macintosh OS/X, and MS/Windows. The aim is to provide an intuitive interface that takes you through the basic steps of data mining, as well as illustrating the R code that is used to achieve this. Whilst the tool itself may be sufficient for all of a user's needs, it also provides a stepping stone to more sophisticated processing and modelling in R itself, for sophisticated and unconstrained data mining.
Rattle is in daily use by Australia's largest team of data miners and by a variety of government and commercial enterprises, world wide. A number of international consultants also use Rattle in their daily business. Users include the Australian Taxation Office, Australian Department of Immigration, Ulster Bank, Toyota Australia, US Geological Survey, Carat Media Network, Institute of Infection and Immunity of the University Hospital of Wales, US National Institutes of Health, AIMIA Loyalty Marketing, Added Value, and many more. It is or has been used for teaching by the McMaster University, Australian National University, University of Canberra, University of Technology Sydney, Yale University, University of Southern Queensland, Revolution Analytics, Habin Institute of Technology Graduate School Shenzhen, and Many more.
The author of Rattle received a 2007 Australia Day Medallion, presented by the Commissioner of Taxation, for leadership and mentoring in Data Mining in the Australian Taxation Office and in Australia, and particularly cited the development and sharing of the Rattle system.
Rattle is also used to teach the practise of data mining. It was the primary tool of instruction for a Data Mining Workshop in Canberra, and at Harbin Institute of Technology, Shenzhen Graduate School (2006), and for the Australian National University's course on Data Mining (since 2006), University of Canberra (since 2010) and University of South Australia (since 2009). It has been used in courses at Yale University, University of Liège Belgium (since 2011), University of Wollongong (since 2010), University of Southern Queensland (since 2010), Australian Consortium for Social and Political Research (2011), University of Technology, Sydney (since 2012), Revolution Analytics (since 2012) and many others.
Further data mining resources are also available from Togaware.
Rattle is open source and freely available from Togaware. You can download Rattle and get familiar with its functionality without any obligation, except for the obligation to freely share! Organisations are also welcome to purchase Rattle, including support for installation and initial training, and ongoing data mining support. Email rattle@togaware.com for details.
Citation
If you use Rattle and report on using it please cite it according to citation("rattle"). You might also reference one of the following:Graham Williams (2011). Data Mining with Rattle and R: The Art of Excavating Data for Knowledge Discovery, Springer, Use R!.
or
Graham Williams (2009). Rattle: A Data Mining GUI for R, Graham J Williams, The R Journal, 1(2):45-55.
Discussion Group and Suggestions
The Rattle Users mailing list is hosted by Google Groups. Questions and suggestions can be posted there. You can [visit the discussion archive] or subscribe by supplying your email address below and clicking the Subscribe button.
Subscribe to:
Posts (Atom)