Difference between revisions of "Machine learning"

From Lesswrongwiki
Jump to: navigation, search
Line 1: Line 1:
 
{{wikilink}}
 
{{wikilink}}
'''Machine Learning''' refers to the general field of study that deals with automated statistical learning and pattern detection by non-biological systems. It is a sub-domain of General Artificial Intelligence that specifically deals with data analysis, modeling and prediction through the knowledge extracted from the previous (training) samples.
+
'''Machine Learning''' refers to the general field of study that deals with automated statistical learning and pattern detection by non-biological systems. It is a sub-domain of General Artificial Intelligence that specifically deals with data analysis, modeling and prediction through the knowledge extracted from the previous (training) samples. As a multi-disciplinary area, it has borrowed concepts and ideas ranging from pure mathematics to cognitive science, all the while trying to exhaustively describe learning systems.
  
 
== Most common algorithms ==
 
== Most common algorithms ==
When considering the most used algorithms in Machine Learning, it's possible to create a simple taxonomy based on their most prominent characteristics.  
+
When considering the most used algorithms in Machine Learning, we can take on multiple approaches to describe its subdivisions. It's possible to create a simple taxonomy based on their most prominent characteristics, such as the following.  
  
The most widely distinction is usually made between the generative (k-NN, k-means clustering, …) vs discriminative (SVM, LDA, …) ones – while the first is able to spontaneously (in a unsupervised way) generate different categories based purely on the data structure, the second kind is only able of distinguishing previously learned classes.
+
Firstly, the most widely distinction is usually made between the generative (k-NN, k-means clustering, …) vs discriminative (Support Vector Machines, Linear Discriminant Analysis, …) ones – while the first is able to spontaneously (in a unsupervised way) generate different categories based purely on the data structure, the second kind is only able of distinguishing previously learned classes (through the feeding of correctly identified data).
  
These same algorithms can be static (such as Neural Networks), disregarding the temporal\sequential characteristics of the data, or dynamic (Hidden Markov Chains, for instance), able to account for those temporal dynamics.
+
These same algorithms can be seen as static (such as Neural Networks), disregarding the temporal\sequential characteristics of the data, or dynamic (Hidden Markov Chains, for instance), able to account for those temporal dynamics and treating time series.
  
The third and last big difference refers to its sensitivity to the variance within the data, that is, the algorithm’s ability to model the training data very closely or more losely (which can actually by a problem commonly referred to as overfitting).
+
The third and last big difference refers to its sensitivity to the variance within the data, that is, the algorithm’s ability to model the training data - either very closely or more losely (which in turn influences its ability to generalize).
  
 
== Applications examples ==
 
== Applications examples ==

Revision as of 09:47, 13 September 2012

Smallwikipedialogo.png
Wikipedia has an article about

Machine Learning refers to the general field of study that deals with automated statistical learning and pattern detection by non-biological systems. It is a sub-domain of General Artificial Intelligence that specifically deals with data analysis, modeling and prediction through the knowledge extracted from the previous (training) samples. As a multi-disciplinary area, it has borrowed concepts and ideas ranging from pure mathematics to cognitive science, all the while trying to exhaustively describe learning systems.

Most common algorithms

When considering the most used algorithms in Machine Learning, we can take on multiple approaches to describe its subdivisions. It's possible to create a simple taxonomy based on their most prominent characteristics, such as the following.

Firstly, the most widely distinction is usually made between the generative (k-NN, k-means clustering, …) vs discriminative (Support Vector Machines, Linear Discriminant Analysis, …) ones – while the first is able to spontaneously (in a unsupervised way) generate different categories based purely on the data structure, the second kind is only able of distinguishing previously learned classes (through the feeding of correctly identified data).

These same algorithms can be seen as static (such as Neural Networks), disregarding the temporal\sequential characteristics of the data, or dynamic (Hidden Markov Chains, for instance), able to account for those temporal dynamics and treating time series.

The third and last big difference refers to its sensitivity to the variance within the data, that is, the algorithm’s ability to model the training data - either very closely or more losely (which in turn influences its ability to generalize).

Applications examples

The use of Machine Learning has been widespread since it’s formal definition in the 50’s. The ability to explore the data structure and make predictions based on previous behavior has been extensively used in areas such as market analysis, natural language processing or even brain-computer interfaces. Amazon’s titles suggestion, for instance, is an example of a deep and recursive system for modeling previous buys and generating possible hypothesis from that data.

Further Reading & References

See Also