"Open source tools have recently reached a level of maturity which makes them suitable for building large-scale real-world systems. At the same time, the field of machine learning has developed a large body of powerful learning algorithms for a wide range of applications. Inspired by similar efforts in bioinformatics (BOSC) or statistics (useR), our aim is to build a forum for open source software in machine learning."

The indispensable reference to getting help without annoying people who might help you.

Gladwell is right when he notes that the 10,000 hour rule keeps appearing as a necessary condition for exceptional performance in many fields. But it’s not sufficient. As Ericsson, along with his colleague Andreas Lehmann, noted in an exceptional overview of this topic, “the mere number of years of experience with relevant activities in a domain is typically only weakly related to performance.”

...

To understand what else is necessary, I’ll turn your attention to a fascinating 2005 study on chess players, published in the journal Applied Cognitive Psychology. After interviewing two large samples of chess players of varied skill, the paper’s authors found that “serious study“ — the arduous task of reviewing past games of better players, trying to predict each move in advance — was the strongest predictor of chess skill.

magic as pedagogy of science. Christopher Brookmyre without the violence.

"We are trying to do for machine learning what Image Magick has done for image processing. That is, we try to keep it simple by providing a big range of functionality in the form of easy-to-use command-line tools. You are free to mix tools from other packages in with ours. You are free to mix your own custom tools with ours. You are free to write scripts that automate your experiments. You are free to use our tools in ways that we could not foresee."

flashcards on the go!

includes lots of tasty goodies radiating out from the SVM core, such as generalised markovian model learning, HDF5, string kernels, multi-language bindings and an emphasis on scalability.

The uniform uncertainty principle and compressed sensing

Best explanation of this sparse image representation thing that I have yet seen. Well wicked.

More high-falutin take on learned helplessness: "I felt silly, because I, too, could have reasoned that out. The water wasn’t draining -- therefore, perhaps the drain was clogged. Basic rationality in action.[1]

But before giving it even ten seconds’ thought, I’d classified the problem as a “mechanical thing”. And I’d remembered I “didn’t know how mechanical things worked” (a cached thought). And then -- prompted by my cached belief that there was a magical “way mechanical things work” that some knew and I didn’t -- I stopped trying to think at all.

“Mechanical things” was for me a mental stopsign -- a blank domain that stayed blank, because I never asked the obvious next questions (questions like “does the dishwasher look unusual in any way? Why is there water at the bottom?”).

it is not enough to have a language for specifying your prior structural beliefs—instead we must have a language for such which results in computationally tractable solutions.

mapreduce alternative thingy for distributed nodes

"And while it may seem crass and anti-intellectual to consider a financial measure of success, it is worth noting that the intellectual offspring of Shannon's theory create several trillion dollars of revenue each year, while the offspring of Chomsky's theories generate well under a billion."

mlpy is a high-performance Python package for predictive modeling. It makes extensive use of NumPy (http://scipy.org) to provide fast N-dimensional array manipulation and easy integration of C code. mlpy provides high level procedures that support, with few lines of code, the design of rich Data Analysis Protocols (DAPs) for preprocessing, clustering, predictive classification, regression and feature selection. Methods are available for feature weighting and ranking, data resampling, error evaluation and experiment landscaping. The package includes tools to measure stability in sets of ranked feature lists.

Markov Logic Networks (MLNs) is a framework that combines statistical and logical reasoning; they have been applied to many data intensive problems including information extraction, entity resolution, text mining, and natural language processing....
...The current version (0.2) of Tuffy is capable of the following MLN tasks:
MAP inference, where we want to find out the most likely possible world;Marginal inference, where we want to estimate marginal probabilities;Weight learning, where we want to learn the weights of MLN rules given training data.Furthermore, Tuffy provides the following functionalities beyond the realm of MLNs:
Datalog: In addition to MLN rules, you can also execute Datalog rules in Tuffy.Functions: Tuffy comes with a library of common numeric/string/boolean functions, which can be used inside an MLN rule. In particular, you can perform arithmetic manipulation and comparison in MLN rules.

what it says

Capo is a revolutionary tool that helps you learn the music in your iTunes library. By slowing your music, and presenting a detailed spectrogram, Capo lets you hear and see your music like never before.

The Clever Algorithms project is an effort to describe a large number of algorithmic techniques from the the field of Artificial Intelligence in a complete, consistent, and centralized manner such that they are accessible, usable, and understandable.

The Clever Algorithms project is currently hosted online at http://www.CleverAlgorithms.com

This is a repository for the Clever Algorithms: Nature-Inspired Programming Recipes project for hosting, versioning, and collaboration, used during the development of the content for deliverables (website and book), as well as for the ongoing maintenance of the outcomes from the project.

machine learning in python, courtesy NICTA again

destigmatizing failure for fun and profit

"k-means clustering, hierarchical clustering and self-organizing maps"

designed for genetic data, but no reason to stick to that. perl, python, c, c++ bindings.

alternative take on the genetic search technique designed to mimic the behaviours of improvising musicians rather than the success and failues of genotypes. Sounds interesting both as a technique and a model.

All the buzzwords I do at once.

coding for semi intelligent agents for the "non programmer".

it's the page of that trendy work-out-your-laws-for-you package

The Agent-based computational Economics mothership is truly here. Tesfatsion ♥♥♥

teh stackexchange mega hub, whee the diaspora of questionably integrate sites that look a bit like the amazing stackoverflow.com site hang out and share expertise.

another Q&A site. unlike stackexchange it doesn't artificially segment communities but lets topics organically grow. unlike vark.com it emphasises collaboration and reputation, and integrates a bit better with social media. And there are lots of nerds i know on it. interesting.

extracting nerds from their shells

"For protein structure prediction, the eventual goal is to have human folders work on proteins that do not have a known structure. This would require first attracting the attention of scientists and biotech companies and convincing them that the process is effective. Another goal is to take folding strategies that  human players have come up with while playing the game, and automate these strategies to make protein-prediction software more effective. These two goals are more or less independent and either or both may happen."

interesting points in here about how to learn to learn to be efficient.

learning in the game-theoretic statistical framework of vovk and shafer

think of it as "direct strategies".

A distributed network for global and local development of generic infrastructures incrementally developed by communities.

A global platform to investigate the new loop of open content, software and hardware for community applications, bringing people together with new technologies and distributed connectivity, unlike the dominant focus of IT industry on security, surveillance and monopoly of information and infrastructures.

bayesian modelling textbook is online

"J-Lab helps journalists and citizens use digital technologies to develop new ways for people to participate in public life with projects on innovations in journalism, citizen media, news games, interactive stories, entrepreneurship, research, training, and publications."

"Peach is a pure-python module, based on SciPy and NumPy to implement algorithms for computational intelligence and machine learning. Methods implemented include, but are not limited to, artificial neural networks, fuzzy logic, genetic algorithms, swarm intelligence and much more."

how to remember your cognitive biases, in musical form

unguided exploration apparently is lame. Is this pedagogic point new only to me?

a sketch of the process of learning the human face manifold by google

a sketch of the process of learning the human face manifold by google

creative-commons remixable, print-on-demand academic textbooks. So device neutral that they even handle physical books.

"What I’d most like to do next is spend a year working full-time on the Software Carpentry course—of all the things I’ve done, it’s the one that I think has the most potential to make scientists’ lives better. My goal is to raise approximately CDN$25,000 from each of half a dozen sponsors so that I can reorganize and revamp the content, add screencasts and video lectures, and generally drag it into the 21st Century. An abbreviated proposal is included below the cut—if you or anyone you know would be interested in discussing possibilities, please give me a shout."

awesome analog hack: an AI that learns noughts-and-crosses, made out of matchboxes and beans

1 dimension exhaustive clustering

Pesky Pesce on learning 2.0-wise

5 easy pieces to practice your code. Except that its actually thousands of easy pieces. and some of them are in fact wacky logical puzzles. (warning - there may be subliminal ads in here for haskell)

Megan seems to have some interest in getting the education internetted

stack overflow, but for maths. There is support for answer bounties, supposedly. (Millennium prize lite?)

Instant guid to getting started with machine learning

What's wrong with blackboard (pt 1)

kind of an equivalent of LETS for learning, online. swap lessons for lessons

There are two ways to have a fast learning algorithm: (a) start with a slow algorithm and speed it up, or (b) build an intrinsically fast learning algorithm. This project is about approach (b), and it's reached a state where it may be useful to others as a platform for research and experimentation.

There are two algorithms, one implementing specialist gradient descent (GD) on squared loss and the other implementing specialist exponentiated gradient descent (SEG) on squared loss.

collective decision making considered as an engineering problem.

the editiable wiki of MIT's collective intelligence thing.

social question answering, where you trade answering questions for answers to questions. fast and interacts with mobile devices, IM, etc, and found me the best pastries in Adelaide.

amazing - a javascript -based neural network doing OCR in the browser. facebook email extraction, anyone?

one of the learning2.0 a-list blogs, blogging about learning.

more python SOM action

image classification using SOM in python

I wonder if i can get a similarity measure (in the informal sense) out of this?

if you'd not run into this particular handy toolkit - machine learning for python!

"PyBrain is a modular Machine Learning Library for Python. It's goal is to offer flexible, easy-to-use yet still powerful algorithms for Machine Learning Tasks and a variety of predefined environments to test and compare your algorithms."

build your own complex adaptive problem solving agents, they reckon.

Awesome bunch of educational videos about disaster prevention using puppets. They haven't put this stuff online for free on their own site for some reason, but you can find them on youtube anyway.

open access meets open source meets biology. hack an organism today.

look! learning processes!

in java they don't learn about the dangers of earthquakes; in australia it's floods. The states apparently, is not so good with hurricanes.

a whole bunch of machine learning tutes from some folks who went on to form a google campus

crazy - prject gutenberg, but for children's book. think pictures and sound, all online.

alex steffen's (long distance) take on what disaster recovery could be

What could happen to cause an entire indonesian village to be replaced with alien bubble domes? ("When earthquakes become high school science projects")

why is twitter such a fertile ground of visualisation?

fascinating account of learning via some brief diversions into the methodology of social science

wow. i can't believe anyone had the guts to create this - a standard to exchange what are are interested in across different platforms. *damn* my software early-adopter persona, I'm going to piss away hours on this one.

a moodle fan talks web2.0 to the educational sector