[ Skip to the content ]

Institute of Formal and Applied Linguistics Wiki

[ Back to the navigation ]

You are here: start » courses » rg » 2014 » perceptron

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Next revision 		Previous revision
courses:rg:2014:perceptron [2014/10/13 22:43]
ufal created 		courses:rg:2014:perceptron [2014/10/14 00:28] (current)
ufal
Line 1: Line 1:
-**Questions are not final yet and may change slightly.**+===== Paper =====
  
 +
 +Michael Collins: [[http://ucrel.lancs.ac.uk/acl/W/W02/W02-1001.pdf|Discriminative Training Methods for Hidden Markov Models: Theory and Experiments with Perceptron Algorithms]], EMNLP 2002. 
 +
 +===== Questions =====
  
 1. Suppose you have a tagset consisting of two tags, N(noun), X(not noun) and a training sentence: 1. Suppose you have a tagset consisting of two tags, N(noun), X(not noun) and a training sentence:
Line 6: Line 10:
 During the training, this best tag sequence for this sentence is found: During the training, this best tag sequence for this sentence is found:
     N N X N X     N N X N X
-How would this result alter values of alpha_X,X,X, alpha_N,father?+How would this result alter values of //α<sub>X,X,X</sub>// and //α<sub>N,father</sub>//<nowiki>?</nowiki>
 Supposing that the best tag sequence won't change, what would be your answer if "father/N" is replaced by "Luke/X"? Supposing that the best tag sequence won't change, what would be your answer if "father/N" is replaced by "Luke/X"?
  
 2. Suppose this tagged sentence as the only entry in your training data: 2. Suppose this tagged sentence as the only entry in your training data:
   a/DT boy/NN saw/VBD a/DT girl/NN with/IN a/DT nice/JJ hat/NN   a/DT boy/NN saw/VBD a/DT girl/NN with/IN a/DT nice/JJ hat/NN
-  How many features will a tagger from section 2.4 have, when its training is identical to the one from section 2.1?+How many features will a tagger from section 2.4 have, when its training is identical to the one from section 2.1?
 (For some reasons, you want to use all 36 tags from Penn Treebank tagset.) (For some reasons, you want to use all 36 tags from Penn Treebank tagset.)
-Which tag sequence (''z_[1:n_1]'') will be selected for this first sentence in the first iteration of algorithm?+Which tag sequence (z<sub>[1:n_1]</sub>) will be selected for this first sentence in the first iteration of algorithm?
  
 3. What is the difference between Maximum-Entropy and the Perceptron model training in the experiments? 3. What is the difference between Maximum-Entropy and the Perceptron model training in the experiments?

[ Back to the navigation ] [ Back to the content ]