Mercurial > pylearn
comparison learner.py @ 14:5ede27026e05
Working on gradient_based_learner
author | bengioy@bengiomac.local |
---|---|
date | Wed, 26 Mar 2008 22:56:13 -0400 |
parents | 633453635d51 |
children | 266c68cb6136 |
comparison
equal
deleted
inserted
replaced
13:633453635d51 | 14:5ede27026e05 |
---|---|
21 learner has a random number generator that is set to use a | 21 learner has a random number generator that is set to use a |
22 different seed each time it forget() is called. | 22 different seed each time it forget() is called. |
23 """ | 23 """ |
24 raise NotImplementedError | 24 raise NotImplementedError |
25 | 25 |
26 def update(self,training_set): | 26 def update(self,training_set,train_stats_collector=None): |
27 """ | 27 """ |
28 Continue training a learner, with the evidence provided by the given training set. | 28 Continue training a learner, with the evidence provided by the given training set. |
29 Hence update can be called multiple times. This is particularly useful in the | 29 Hence update can be called multiple times. This is particularly useful in the |
30 on-line setting or the sequential (Bayesian or not) settings. | 30 on-line setting or the sequential (Bayesian or not) settings. |
31 The result is a function that can be applied on data, with the same | 31 The result is a function that can be applied on data, with the same |
32 semantics of the Learner.use method. | 32 semantics of the Learner.use method. |
33 The user may optionally provide a training StatsCollector that is used to record | |
34 some statistics of the outputs computed during training. | |
33 """ | 35 """ |
34 return self.use # default behavior is 'non-adaptive', i.e. update does not do anything | 36 return self.use # default behavior is 'non-adaptive', i.e. update does not do anything |
35 | 37 |
36 | 38 |
37 def __call__(self,training_set): | 39 def __call__(self,training_set): |