Mercurial > pylearn

comparison linear_regression.py @ 421:e01f17be270a

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Kernel regression learning algorithm
author Yoshua Bengio <bengioy@iro.umontreal.ca>
date Sat, 19 Jul 2008 10:11:22 -0400
parents 43d9aa93934e
children fa4a5fee53ce
comparison
equal deleted inserted replaced
420:040cb796f4e0 421:e01f17be270a
32 The training_set must have fields "input" and "target". 32 The training_set must have fields "input" and "target".
33 The test_set must have field "input", and needs "target" if 33 The test_set must have field "input", and needs "target" if
34 we want to compute the squared errors. 34 we want to compute the squared errors.
35 35
36 The predictor parameters are obtained analytically from the training set. 36 The predictor parameters are obtained analytically from the training set.
37
38 *** NOT IMPLEMENTED YET ***
37 Training can proceed sequentially (with multiple calls to update with 39 Training can proceed sequentially (with multiple calls to update with
38 different disjoint subsets of the training sets). After each call to 40 different disjoint subsets of the training sets). After each call to
39 update the predictor is ready to be used (and optimized for the union 41 update the predictor is ready to be used (and optimized for the union
40 of all the training sets passed to update since construction or since 42 of all the training sets passed to update since construction or since
41 the last call to forget). 43 the last call to forget).
42 44 ***************************
45
43 For each (input[t],output[t]) pair in a minibatch,:: 46 For each (input[t],output[t]) pair in a minibatch,::
44 47
45 output_t = b + W * input_t 48 output_t = b + W * input_t
46 49
47 where b and W are obtained by minimizing:: 50 where b and W are obtained by minimizing::
72 = example-wise squared error 75 = example-wise squared error
73 """ 76 """
74 def __init__(self, L2_regularizer=0,minibatch_size=10000): 77 def __init__(self, L2_regularizer=0,minibatch_size=10000):
75 self.L2_regularizer=L2_regularizer 78 self.L2_regularizer=L2_regularizer
76 self.equations = LinearRegressionEquations() 79 self.equations = LinearRegressionEquations()
77 self.minibatch_size=1000 80 self.minibatch_size=minibatch_size
78 81
79 def __call__(self,trainset): 82 def __call__(self,trainset):
80 first_example = trainset[0] 83 first_example = trainset[0]
81 n_inputs = first_example['input'].size 84 n_inputs = first_example['input'].size
82 n_outputs = first_example['target'].size 85 n_outputs = first_example['target'].size