Mercurial > pylearn
annotate learner.py @ 78:3499918faa9d
In the middle of designing TLearner
author | bengioy@bengiomac.local |
---|---|
date | Mon, 05 May 2008 09:35:30 -0400 |
parents | 1e2bb5bad636 |
children | c4726e19b8ec |
rev | line source |
---|---|
1
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
1 |
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
2 from dataset import * |
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
3 |
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
4 class Learner(object): |
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
5 """Base class for learning algorithms, provides an interface |
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
6 that allows various algorithms to be applicable to generic learning |
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
7 algorithms. |
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
8 |
10
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
9 A Learner can be seen as a learning algorithm, a function that when |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
10 applied to training data returns a learned function, an object that |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
11 can be applied to other data and return some output data. |
1
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
12 """ |
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
13 |
10
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
14 def __init__(self): |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
15 pass |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
16 |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
17 def forget(self): |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
18 """ |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
19 Reset the state of the learner to a blank slate, before seeing |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
20 training data. The operation may be non-deterministic if the |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
21 learner has a random number generator that is set to use a |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
22 different seed each time it forget() is called. |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
23 """ |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
24 raise NotImplementedError |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
25 |
14 | 26 def update(self,training_set,train_stats_collector=None): |
10
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
27 """ |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
28 Continue training a learner, with the evidence provided by the given training set. |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
29 Hence update can be called multiple times. This is particularly useful in the |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
30 on-line setting or the sequential (Bayesian or not) settings. |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
31 The result is a function that can be applied on data, with the same |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
32 semantics of the Learner.use method. |
75
90e4c0784d6e
Added draft of LinearRegression learner
bengioy@bengiomac.local
parents:
20
diff
changeset
|
33 |
14 | 34 The user may optionally provide a training StatsCollector that is used to record |
75
90e4c0784d6e
Added draft of LinearRegression learner
bengioy@bengiomac.local
parents:
20
diff
changeset
|
35 some statistics of the outputs computed during training. It is update(d) during |
90e4c0784d6e
Added draft of LinearRegression learner
bengioy@bengiomac.local
parents:
20
diff
changeset
|
36 training. |
10
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
37 """ |
13
633453635d51
Starting to work on gradient_based_learner.py
bengioy@bengiomac.local
parents:
10
diff
changeset
|
38 return self.use # default behavior is 'non-adaptive', i.e. update does not do anything |
10
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
39 |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
40 |
20
266c68cb6136
Minor editions, plus adding untested ApplyFunctionDataset for GradientLearner in the works.
bengioy@bengiomac.local
parents:
14
diff
changeset
|
41 def __call__(self,training_set,train_stats_collector=None): |
10
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
42 """ |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
43 Train a learner from scratch using the provided training set, |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
44 and return the learned function. |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
45 """ |
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
46 self.forget() |
20
266c68cb6136
Minor editions, plus adding untested ApplyFunctionDataset for GradientLearner in the works.
bengioy@bengiomac.local
parents:
14
diff
changeset
|
47 return self.update(learning_task,train_stats_collector) |
1
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
48 |
20
266c68cb6136
Minor editions, plus adding untested ApplyFunctionDataset for GradientLearner in the works.
bengioy@bengiomac.local
parents:
14
diff
changeset
|
49 def use(self,input_dataset,output_fields=None,copy_inputs=True): |
10
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
50 """Once a Learner has been trained by one or more call to 'update', it can |
1
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
51 be used with one or more calls to 'use'. The argument is a DataSet (possibly |
10
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
52 containing a single example) and the result is a DataSet of the same length. |
1
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
53 If output_fields is specified, it may be use to indicate which fields should |
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
54 be constructed in the output DataSet (for example ['output','classification_error']). |
20
266c68cb6136
Minor editions, plus adding untested ApplyFunctionDataset for GradientLearner in the works.
bengioy@bengiomac.local
parents:
14
diff
changeset
|
55 Optionally, if copy_inputs, the input fields (of the input_dataset) can be made |
75
90e4c0784d6e
Added draft of LinearRegression learner
bengioy@bengiomac.local
parents:
20
diff
changeset
|
56 visible in the output DataSet returned by this method. |
10
80bf5492e571
Rewrote learner.py according to the specs in the wiki for learners.
bengioy@esprit.iro.umontreal.ca
parents:
1
diff
changeset
|
57 """ |
1
2cd82666b9a7
Added statscollector and started writing dataset and learner.
bengioy@esprit.iro.umontreal.ca
parents:
0
diff
changeset
|
58 raise NotImplementedError |
77
1e2bb5bad636
toying with different ways to implement learners
bengioy@bengiomac.local
parents:
75
diff
changeset
|
59 |
78 | 60 def attribute_names(self): |
61 """ | |
62 A Learner may have attributes that it wishes to export to other objects. To automate | |
63 such export, sub-classes should define here the names (list of strings) of these attributes. | |
64 """ | |
65 return [] | |
77
1e2bb5bad636
toying with different ways to implement learners
bengioy@bengiomac.local
parents:
75
diff
changeset
|
66 |
78 | 67 class TLearner(Learner): |
68 """ | |
69 TLearner is a virtual class of Learners that attempts to factor out of the definition | |
70 of a learner the steps that are common to many implementations of learning algorithms, | |
71 so as to leave only "the equations" to define in particular sub-classes, using Theano. | |
72 | |
73 In the default implementations of use and update, it is assumed that the 'use' and 'update' methods | |
74 visit examples in the input dataset sequentially. In the 'use' method only one pass through the dataset is done, | |
75 whereas the sub-learner may wish to iterate over the examples multiple times. Subclasses where this | |
76 basic model is not appropriate can simply redefine update or use. | |
77 | |
78 Sub-classes must provide the following functions and functionalities: | |
79 - attributeNames(): defines all the names of attributes which can be used as fields or | |
80 attributes in input/output datasets or in stats collectors. | |
81 All these attributes are expected to be theano.Result objects | |
82 (with a .data property and recognized by theano.Function for compilation). | |
83 The sub-class constructor defines the relations between | |
84 the Theano variables that may be used by 'use' and 'update' | |
85 or by a stats collector. | |
86 - defaultOutputFields(input_fields): return a list of default dataset output fields when | |
87 None are provided by the caller of use. | |
88 - | |
89 | |
90 """ | |
91 |