Comments on: Machine Learning: Classification http://www.grok.in/blog/2008/03/27/machine-learning-classification/ (ignorance killed the cat, curiosity was framed) Wed, 16 Jun 2010 18:48:41 +0000 hourly 1 By: Will Dwinnell http://www.grok.in/blog/2008/03/27/machine-learning-classification/comment-page-1/#comment-96 Will Dwinnell Sat, 05 Apr 2008 09:30:51 +0000 http://www.grok.in/blog/2008/03/27/machine-learning-classification/#comment-96 In the particular case of spam detectors, one very common input data representation is called "bag of words", which is just a list of the distinct words in the message, or a 0/1 vector representing the same. Some pre-processing will likely precede the construction of a bag of words, such as removal of very common words ("the", "of"), de-stemming ("walking", "walks" and "walked" become "walk") and selection of words with high predictive value ("Viagra"). In the particular case of spam detectors, one very common input data representation is called “bag of words”, which is just a list of the distinct words in the message, or a 0/1 vector representing the same. Some pre-processing will likely precede the construction of a bag of words, such as removal of very common words (“the”, “of”), de-stemming (“walking”, “walks” and “walked” become “walk”) and selection of words with high predictive value (“Viagra”).

]]> By: Siddhartha Reddy http://www.grok.in/blog/2008/03/27/machine-learning-classification/comment-page-1/#comment-61 Siddhartha Reddy Sun, 30 Mar 2008 04:22:05 +0000 http://www.grok.in/blog/2008/03/27/machine-learning-classification/#comment-61 @sanket, The training data is made up of entities that need to be classified. So, in the case of the spam classifier, the training data is emails labeled as 'spam' and 'not-spam'. Words/sets-of-words are used as 'features' for the classification process. What this means is that the classification is made based on the occurrence of words. Feature selection is an important and usually non-trivial step in classification. I'll be posting about this later on. @sanket,

The training data is made up of entities that need to be classified. So, in the case of the spam classifier, the training data is emails labeled as ’spam’ and ‘not-spam’.

Words/sets-of-words are used as ‘features’ for the classification process. What this means is that the classification is made based on the occurrence of words.

Feature selection is an important and usually non-trivial step in classification. I’ll be posting about this later on.

]]> By: sanket http://www.grok.in/blog/2008/03/27/machine-learning-classification/comment-page-1/#comment-58 sanket Sat, 29 Mar 2008 18:41:59 +0000 http://www.grok.in/blog/2008/03/27/machine-learning-classification/#comment-58 But what is the training data in this case? Words? Sets of words? But what is the training data in this case? Words? Sets of words?

]]>