www.groksolutions.com – based on NuPIC (Numenta Platform for Intelligent Computiong).
I am particularly interested in unsupervised, brain inspired learning algorithms such as Self Organizing Maps (SOM), Cortical Learning Algorithm (CLA), and Hierarchical Temporal Memory (HTM). Those that are attempting to model temporal and (or) spatial component.
Since J. Hawkins released NuPIC as on Open Source, I have been studying and trying to apply with new data sets.
Source Code can be downloaded from:
www.numenta.org
Similarly, based on Hierarchical Temporal Memory and CLA concepts there is another open source project OpenHTM:
http://sourceforge.net/projects/openhtm/
These projects open a number of research questions and opportunities.
The next big thing is in Machine Learning can be unsupervised algorithms and feature extraction (dimension reduction) with very small loss of information, prediction of lot of variables with high occuracy using a few input
EyeQuant: http://eyequant.com/science - predict human attention when looking at websites based on eye-tracking studies; they use algorithms that analyze images based on statistical feature sets that are known to drive human attention in the first few seconds – effectively a simulation of what the human visual cortex does when it tries to figure out what is most ’interesting’ about a scene.