Mercurial > pylearn

comparison doc/v2_planning/use_cases.txt @ 1097:8be7928cc1aa

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
use_cases: Added couple questions
author Olivier Delalleau <delallea@iro>
date Mon, 13 Sep 2010 09:38:49 -0400
parents a65598681620
children b422cbaddc52
comparison
equal deleted inserted replaced
1096:2bbc294fa5ac 1097:8be7928cc1aa
57 There are many ways that the training could be configured, but here is one: 57 There are many ways that the training could be configured, but here is one:
58 58
59 59
60 vm.call( 60 vm.call(
61 halflife_stopper( 61 halflife_stopper(
62 # OD: is n_hidden supposed to be n_classes instead?
62 initial_model=random_linear_classifier(MNIST.n_inputs, MNIST.n_hidden, r_seed=234432), 63 initial_model=random_linear_classifier(MNIST.n_inputs, MNIST.n_hidden, r_seed=234432),
63 burnin=100, 64 burnin=100,
64 score_fn = vm_lambda(('learner_obj',), 65 score_fn = vm_lambda(('learner_obj',),
65 classification_accuracy( 66 classification_accuracy(
66 examples=MNIST.validation_dataset, 67 examples=MNIST.validation_dataset,
98 99
99 K-fold cross validation of a classifier 100 K-fold cross validation of a classifier
100 --------------------------------------- 101 ---------------------------------------
101 102
102 splits = kfold_cross_validate( 103 splits = kfold_cross_validate(
104 # OD: What would these parameters mean?
103 indexlist = range(1000) 105 indexlist = range(1000)
104 train = 8, 106 train = 8,
105 valid = 1, 107 valid = 1,
106 test = 1, 108 test = 1,
107 ) 109 )