Revision history[back]
click to hide/show revision 1
Revision n. 1

Apr 28 '11 at 00:28

Jacob%20Jensen's gravatar image

Jacob Jensen
1899315562

What are some Novel, Emerging Areas of Machine Learning

What are novel research areas and subareas of machine learning whose conceptual birth was in the last 5 years or so?

A significant portion of machine learning research could be called "collaboration between the past and present". ML researchers are taking and re-purposing models from math, statistics, physics and old machine learning research that are already well-understood in those fields and finding new applications and extensions.

However, I want to know: what are some areas of Machine Learning that are currently quite active that have had their mathematical or conceptual foundations either discovered or greatly expanded upon in the last few years, either in machine learning or an outside but related discipline?

I'm looking less for meta-applications (multi-task learning, semi-supervised learning) and more for methods that have a strong mathematical basis.

click to hide/show revision 2
Revision n. 2

Apr 28 '11 at 00:29

Jacob%20Jensen's gravatar image

Jacob Jensen
1899315562

What are some Novel, Emerging Areas of Machine Learning

What are novel research areas and subareas of machine learning whose conceptual birth was in the last 5 years or so?

A significant portion of machine learning research could be called "collaboration between the past and present". ML researchers are taking and re-purposing models from math, statistics, physics and old machine learning research that are already well-understood in those fields and finding new applications and extensions.

However, I want to know: what are some areas of Machine Learning that are currently quite active that have had their mathematical or conceptual foundations either discovered or greatly expanded upon in the last few years, either in machine learning or an outside but related discipline?

I'm looking less for meta-applications (multi-task learning, semi-supervised learning) and more for methods that have a strong mathematical basis.

powered by OSQA

User submitted content is under Creative Commons: Attribution - Share Alike; Other things copyright (C) 2010, MetaOptimize LLC.