Mercurial > pylearn

annotate doc/v2_planning/optimization.txt @ 1153:ae5ba6206fd3

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
a first draft of pseudo-code for logreg .. using version B (?) approach
author Razvan Pascanu <r.pascanu@gmail.com>
date Thu, 16 Sep 2010 17:34:30 -0400
parents 7c5dc11c850a
children f2105a06201c
rev   line source
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1057
diff changeset
1 =========================
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1057
diff changeset
2 Optimization for Learning
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1057
diff changeset
3 =========================
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1057
diff changeset
4
1068
9fe0f0755b03 optimization: Fixed typo in my name :o
Olivier Delalleau <delallea@iro>
parents: 1064
diff changeset
5 Members: Bergstra, Lamblin, Delalleau, Glorot, Breuleux, Bordes
1064
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1057
diff changeset
6 Leader: Bergstra
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1057
diff changeset
7
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1057
diff changeset
8
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1057
diff changeset
9
a41cc29cee26 v2planning optimization - API draft
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1057
diff changeset
10 Initial Writeup by James
1009
dc5185cca21e Added files for Coding Style and Optimization committees
Olivier Delalleau <delallea@iro>
parents:
diff changeset
11 =========================================
dc5185cca21e Added files for Coding Style and Optimization committees
Olivier Delalleau <delallea@iro>
parents:
diff changeset
12
1013
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
13
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
14
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
15 Previous work - scikits, openopt, scipy provide function optimization
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
16 algorithms. These are not currently GPU-enabled but may be in the future.
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
17
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
18
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
19 IS PREVIOUS WORK SUFFICIENT?
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
20 --------------------------------
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
21
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
22 In many cases it is (I used it for sparse coding, and it was ok).
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
23
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
24 These packages provide batch optimization, whereas we typically need online
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
25 optimization.
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
26
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
27 It can be faster (to run) and more convenient (to implement) to have
1016
618b9fdbfda5 optimization: Minor typo fixes
Olivier Delalleau <delallea@iro>
parents: 1013
diff changeset
28 optimization algorithms as Theano update expressions.
1013
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
29
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
30
1016
618b9fdbfda5 optimization: Minor typo fixes
Olivier Delalleau <delallea@iro>
parents: 1013
diff changeset
31 What optimization algorithms do we want/need?
1013
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
32 ---------------------------------------------
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
33
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
34 - sgd
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
35 - sgd + momentum
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
36 - sgd with annealing schedule
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
37 - TONGA
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
38 - James Marten's Hessian-free
1027
a1b6ccd5b6dc few comments added
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 1016
diff changeset
39 - Conjugate gradients, batch and (large) mini-batch [that is also what Marten's thing does]
1013
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
40
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
41 Do we need anything to make batch algos work better with Pylearn things?
1027
a1b6ccd5b6dc few comments added
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 1016
diff changeset
42 - conjugate methods? yes
a1b6ccd5b6dc few comments added
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 1016
diff changeset
43 - L-BFGS? maybe, when needed
1013
5e9a3d9bc0b4 optimization - added some text
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1009
diff changeset
44
1027
a1b6ccd5b6dc few comments added
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 1016
diff changeset
45
a1b6ccd5b6dc few comments added
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 1016
diff changeset
46
a1b6ccd5b6dc few comments added
Yoshua Bengio <bengioy@iro.umontreal.ca>
parents: 1016
diff changeset
47
1057
baf1988db557 v2planning optimization - added API
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1036
diff changeset
48
baf1988db557 v2planning optimization - added API
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1036
diff changeset
49 Proposal for API
baf1988db557 v2planning optimization - added API
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1036
diff changeset
50 ================
baf1988db557 v2planning optimization - added API
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1036
diff changeset
51
1071
2cf3ad953bf9 optimization: Removed duplicated API draft and asked question
Olivier Delalleau <delallea@iro>
parents: 1068
diff changeset
52 See api_optimization.txt.
1057
baf1988db557 v2planning optimization - added API
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1036
diff changeset
53
1149
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
54 OD asks: Do we really need a different file? If yes, maybe create a subdirectory
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
55 to be able to easily find all files related to optimization?
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
56
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
57 JB replies: Yoshua's orders.
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
58
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
59
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
60
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
61 OD asks: Could it be more convenient for x0 to be a list?
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
62
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
63 JB replies: Yes, but that's not the interface used by other minimize()
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
64 routines (e.g. in scipy). Maybe another list-based interface is required?
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
65
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
66 OD replies: I think most people would prefer to use a list-based interface, so
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
67 they don't have to manually pack / unpack multiple arrrays of parameters. So I
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
68 would vote in favor or having both (where the main reason to also provide a
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
69 non-list interface would be to allow one to easily switch e.g. to scipy's
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
70 minimize).
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
71 I would guess the reason scipy's interface is like this is because it makes
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
72 it easier for the optimization algorithm. However, this does not really
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
73 matter if we are just wrapping a theano-based algorithm (that already has
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
74 to handle multiple parameters), and avoiding useless data copies on each call
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
75 to f / df can only help speed-wise.
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
76 JB replies: Done, I added possibility that x0 is list of ndarrays to the api
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
77 doc.
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
78
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
79
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
80
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
81 OD asks: Why make a difference between iterative and one-shot versions? A one-shot
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
82 algorithm can be seen as an iterative one that stops after its first
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
83 iteration. The difference I see between the two interfaces proposed here
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
84 is mostly that one relies on Theano while the other one does not, but
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
85 hopefully a non-Theano one can be created by simply wrapping around the
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
86 Theano one.
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
87
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
88 JB replies: Right, it would make more sense to distinguish them by the fact that
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
89 one works on Theano objects, and the other on general Python callable functions.
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
90 There is room for an iterative numpy interface, but I didn't make it yet. Would
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
91 that answer your question?
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
92
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
93 OD replies and asks: Partly. Do we really need a non-iterative interface?
7c5dc11c850a cleaning up api_optimization
James Bergstra <bergstrj@iro.umontreal.ca>
parents: 1071
diff changeset
94