Mercurial > pylearn

annotate doc/v2_planning/api_optimization.txt @ 1135:a1957faecc9b

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
revised plugin interface and implementation
author Olivier Breuleux <breuleuo@iro.umontreal.ca>
date Thu, 16 Sep 2010 02:58:24 -0400
parents c5c7ba805a2f
children 7c5dc11c850a
rev   line source
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
1 Optimization API
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
2 ================
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
3
1069
16ea3e5c5a7a api_optimization: Couple questions
Olivier Delalleau <delallea@iro>
parents: 1065
diff changeset
4 Members: Bergstra, Lamblin, Delalleau, Glorot, Breuleux, Bordes
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
5 Leader: Bergstra
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
6
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
7
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
8 Description
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
9 -----------
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
10
1065
2bbc464d6ed0 typo in doc
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1064
diff changeset
11 This API is for iterative optimization algorithms, such as:
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
12
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
13 - stochastic gradient descent (incl. momentum, annealing)
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
14 - delta bar delta
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
15 - conjugate methods
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
16 - L-BFGS
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
17 - "Hessian Free"
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
18 - SGD-QN
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
19 - TONGA
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
20
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
21 The API includes an iterative interface based on Theano, and a one-shot
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
22 interface similar to SciPy and MATLAB that is based on Python and Numpy, that
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
23 only uses Theano for the implementation.
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
24
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
25
1100
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
26 Theano Interface
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
27 -----------------
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
28
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
29 The theano interface to optimization algorithms is to ask for a dictionary of
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
30 updates that can be used in theano.function. Implementations of iterative
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
31 optimization algorithms should be global functions with a signature like
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
32 'iterative_optimizer'.
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
33
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
34 def iterative_optimizer(parameters,
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
35 cost=None,
1100
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
36 gradients=None,
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
37 stop=None,
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
38 updates=None,
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
39 **kwargs):
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
40 """
1100
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
41 :param parameters: list or tuple of Theano variables
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
42 that we want to optimize iteratively. If we're minimizing f(x), then
1100
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
43 together, these variables represent 'x'. Typically these are shared
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
44 variables and their values are the initial values for the minimization
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
45 algorithm.
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
46
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
47 :param cost: scalar-valued Theano variable that computes an exact or noisy estimate of
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
48 cost (what are the conditions on the noise?). Some algorithms might
1100
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
49 need an exact cost, some algorithms might ignore the cost if the
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
50 gradients are given.
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
51
1100
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
52 :param gradients: list or tuple of Theano variables representing the gradients on
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
53 the corresponding parameters. These default to tensor.grad(cost,
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
54 parameters).
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
55
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
56 :param stop: a shared variable (scalar integer) that (if provided) will be
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
57 updated to say when the iterative minimization algorithm has finished
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
58 (1) or requires more iterations (0).
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
59
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
60 :param updates: a dictionary to update with the (var, new_value) items
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
61 associated with the iterative algorithm. The default is a new empty
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
62 dictionary. A KeyError is raised in case of key collisions.
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
63
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
64 :param kwargs: algorithm-dependent arguments
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
65
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
66 :returns: a dictionary mapping each parameter to an expression that it
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
67 should take in order to carry out the optimization procedure.
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
68
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
69 If all the parameters are shared variables, then this dictionary may be
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
70 passed as the ``updates`` argument to theano.function.
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
71
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
72 There may be more key,value pairs in the dictionary corresponding to
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
73 internal variables that are part of the optimization algorithm.
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
74
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
75 """
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
76
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
77
1100
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
78 Numpy Interface
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
79 ---------------
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
80
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
81 The numpy interface to optimization algorithms is supposed to mimick
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
82 scipy's. Its arguments are numpy arrays, and functions that manipulate numpy
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
83 arrays.
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
84
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
85 TODO: There is also room for an iterative object (that doesn't hog program
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
86 control) but which nonetheless works on numpy objects. Actually minimize() should
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
87 use this iterative interface under the hood.
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
88
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
89 def minimize(x0, f, df, opt_algo, **kwargs):
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
90 """
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
91 Return a point x_new that minimizes function `f` with derivative `df`.
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
92
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
93 This is supposed to provide an interface similar to scipy's minimize
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
94 routines, or MATLAB's.
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
95
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
96 :type x0: numpy ndarray
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
97 :param x0: starting point for minimization
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
98
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
99 :type f: python callable mapping something like x0 to a scalar
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
100 :param f: function to minimize
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
101
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
102 :type df: python callable mapping something like x0 to the derivative of f at that point
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
103 :param df: derivative of `f`
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
104
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
105 :param opt_algo: one of the functions that implements the
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
106 `iterative_optimizer` interface.
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
107
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
108 :param kwargs: passed through to `opt_algo`
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
109
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
110 """
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
111
1100
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
112 OD asks: Could it be more convenient for x0 to be a list?
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
113
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
114 JB replies: Yes, but that's not the interface used by other minimize()
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
115 routines (e.g. in scipy). Maybe another list-based interface is required?
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
116
1102
e7c52923f122 api_optimization: Couple replies
Olivier Delalleau <delallea@iro>
parents: 1100
diff changeset
117 OD replies: I think most people would prefer to use a list-based interface, so
e7c52923f122 api_optimization: Couple replies
Olivier Delalleau <delallea@iro>
parents: 1100
diff changeset
118 they don't have to manually pack / unpack multiple arrrays of parameters. So I
e7c52923f122 api_optimization: Couple replies
Olivier Delalleau <delallea@iro>
parents: 1100
diff changeset
119 would vote in favor or having both (where the main reason to also provide a
e7c52923f122 api_optimization: Couple replies
Olivier Delalleau <delallea@iro>
parents: 1100
diff changeset
120 non-list interface would be to allow one to easily switch e.g. to scipy's
e7c52923f122 api_optimization: Couple replies
Olivier Delalleau <delallea@iro>
parents: 1100
diff changeset
121 minimize).
e7c52923f122 api_optimization: Couple replies
Olivier Delalleau <delallea@iro>
parents: 1100
diff changeset
122 I would guess the reason scipy's interface is like this is because it makes
1108
c5c7ba805a2f api_optimization: Edited comment - Unpacking is actually needed on each call to f/df
Olivier Delalleau <delallea@iro>
parents: 1102
diff changeset
123 it easier for the optimization algorithm. However, this does not really
c5c7ba805a2f api_optimization: Edited comment - Unpacking is actually needed on each call to f/df
Olivier Delalleau <delallea@iro>
parents: 1102
diff changeset
124 matter if we are just wrapping a theano-based algorithm (that already has
c5c7ba805a2f api_optimization: Edited comment - Unpacking is actually needed on each call to f/df
Olivier Delalleau <delallea@iro>
parents: 1102
diff changeset
125 to handle multiple parameters), and avoiding useless data copies on each call
c5c7ba805a2f api_optimization: Edited comment - Unpacking is actually needed on each call to f/df
Olivier Delalleau <delallea@iro>
parents: 1102
diff changeset
126 to f / df can only help speed-wise.
1100
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
127
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
128 OD asks: Why make a difference between iterative and one-shot versions? A one-shot
1069
16ea3e5c5a7a api_optimization: Couple questions
Olivier Delalleau <delallea@iro>
parents: 1065
diff changeset
129 algorithm can be seen as an iterative one that stops after its first
16ea3e5c5a7a api_optimization: Couple questions
Olivier Delalleau <delallea@iro>
parents: 1065
diff changeset
130 iteration. The difference I see between the two interfaces proposed here
16ea3e5c5a7a api_optimization: Couple questions
Olivier Delalleau <delallea@iro>
parents: 1065
diff changeset
131 is mostly that one relies on Theano while the other one does not, but
16ea3e5c5a7a api_optimization: Couple questions
Olivier Delalleau <delallea@iro>
parents: 1065
diff changeset
132 hopefully a non-Theano one can be created by simply wrapping around the
16ea3e5c5a7a api_optimization: Couple questions
Olivier Delalleau <delallea@iro>
parents: 1065
diff changeset
133 Theano one.
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents:
diff changeset
134
1100
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
135 JB replies: Right, it would make more sense to distinguish them by the fact that
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
136 one works on Theano objects, and the other on general Python callable functions.
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
137 There is room for an iterative numpy interface, but I didn't make it yet. Would
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
138 that answer your question?
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
139
1102
e7c52923f122 api_optimization: Couple replies
Olivier Delalleau <delallea@iro>
parents: 1100
diff changeset
140 OD replies and asks: Partly. Do we really need a non-iterative interface?
1100
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
141
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
142 Examples
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
143 --------
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
144
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
145
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
146 Simple stochastic gradient descent with extra updates:
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
147
153cf820a975 v2planning - updates to api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1069
diff changeset
148 sgd([p], gradients=[g], updates={a:b}, step_size=.1) will return {a:b, p:p-.1*g}