Mercurial > pylearn
comparison doc/v2_planning/architecture.txt @ 1201:46527ae6db53
architecture: Clarified what I meant about saving the model
| author | Olivier Delalleau <delallea@iro> |
|---|---|
| date | Mon, 20 Sep 2010 17:05:15 -0400 |
| parents | 9ff2242a817b |
| children | b9d0a326e3e7 |
comparison
equal
deleted
inserted
replaced
| 1200:acfd5e747a75 | 1201:46527ae6db53 |
|---|---|
| 159 - Plugins with a global scheduler driving the experiment (Razvan's team) | 159 - Plugins with a global scheduler driving the experiment (Razvan's team) |
| 160 - Objects, with basic hooks at predefined places (Pascal L.'s team) | 160 - Objects, with basic hooks at predefined places (Pascal L.'s team) |
| 161 - Existing objects and code (including dbi and Jobman), with some more | 161 - Existing objects and code (including dbi and Jobman), with some more |
| 162 pieces to tie things together (Fred B.) | 162 pieces to tie things together (Fred B.) |
| 163 | 163 |
| 164 OD comments: We were in a hurry to close the meeting and I did not have time | |
| 165 to really explain what I meant when I suggested we should add the requirement | |
| 166 of saving the final "best" model. What I had in mind is a typical "applied ML" | |
| 167 experiment, i.e. the following approach that hopefully can be understood just | |
| 168 by writing it down in the form of a processing pipeline. The double cross | |
| 169 validation step, whose goal is to obtain an estimate of the generalization | |
| 170 error of our final model, is: | |
| 171 data -> k_fold_outer(preprocessing -> k_fold_inner(dbn -> evaluate) -> select_best -> retrain_on_all_data -> evaluate) -> evaluate | |
| 172 Once this is done, the model we want to save is obtained by doing | |
| 173 data -> preprocessing -> k_fold(dbn -> evaluate) -> select_best -> retrain_on_all_data | |
| 174 and we save | |
| 175 preprocessing -> best_model_selected |