[ Skip to the content ]

Institute of Formal and Applied Linguistics Wiki

[ Back to the navigation ]

You are here: start » user » dusek » vystadial » flect

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision 		Previous revision
Next revision Both sides next revision
user:dusek:vystadial:flect [2012/12/19 10:04]
dusek 		user:dusek:vystadial:flect [2013/01/14 16:15]
dusek
Line 4: Line 4:
 2012-11-27 09:53:07,196 Loaded. Vectorizing... 2012-11-27 09:53:07,196 Loaded. Vectorizing...
 2012-11-27 09:55:51,228 Data shape (652544, 301784) 2012-11-27 09:55:51,228 Data shape (652544, 301784)
-/home/odusek/.local-x86_64/lib/python2.7/site-packages/sklearn/feature_selection/univariate_selection.py:94: RuntimeWarning: divide by zero encountered in divide +/home/odusek/.local-x86_64/lib/python2.7/site-packages/sklearn/feature_selection/univariate_selection.py:94: RuntimeWarning: divide by zero encountered in divide  f = msb / msw
-  f = msb / msw+
 2012-11-27 09:56:28,985 Filt shape (652544, 30183) 2012-11-27 09:56:28,985 Filt shape (652544, 30183)
 2012-11-27 09:56:28,986 Training ... 2012-11-27 09:56:28,986 Training ...
Line 22: Line 21:
  
  
-*** Přeprogramování trénování ***+** Přeprogramování trénování **
  
 2012-12-18 11:14:06,606 TREEX-INFO: Loading data set from data/train.arff.gz... 2012-12-18 11:14:06,606 TREEX-INFO: Loading data set from data/train.arff.gz...
 2012-12-18 11:16:20,214 TREEX-INFO: Preparing data set... 2012-12-18 11:16:20,214 TREEX-INFO: Preparing data set...
 2012-12-18 11:18:04,516 TREEX-INFO: Filtering... 2012-12-18 11:18:04,516 TREEX-INFO: Filtering...
-/home/odusek/.local-x86_64/lib/python2.7/site-packages/sklearn/feature_selection/univariate_selection.py:94: RuntimeWarning: divide by zero encountered in divide +/home/odusek/.local-x86_64/lib/python2.7/site-packages/sklearn/feature_selection/univariate_selection.py:94: RuntimeWarning: divide by zero encountered in divide  f = msb / msw
-  f = msb / msw+
 2012-12-18 11:18:24,378 TREEX-INFO: Training... 2012-12-18 11:18:24,378 TREEX-INFO: Training...
 2012-12-18 17:26:14,641 TREEX-INFO: Training done. 2012-12-18 17:26:14,641 TREEX-INFO: Training done.
Line 35: Line 33:
 2012-12-18 17:27:26,210 TREEX-INFO: Saving model to file runs/train-plain/model.pickle.gz 2012-12-18 17:27:26,210 TREEX-INFO: Saving model to file runs/train-plain/model.pickle.gz
 2012-12-18 17:28:30,627 TREEX-INFO: Model successfully saved. 2012-12-18 17:28:30,627 TREEX-INFO: Model successfully saved.
 +
 +
 +** Trénování na datech vyprodukovaných v pythonu **
 +
 +andromeda2:~/od-playground/test/exp-flect$ ../../src/experiment/train_model.py runs/train-plain_pydata/config.py data/train.arff.gz runs/train-plain_pydata/model.pickle.gz data/dtest.arff.gz runs/train-plain_pydata/classif.arff.gz
 +2013-01-04 13:28:17,539 TREEX-INFO: Loading data set from data/train.arff.gz...
 +2013-01-04 13:29:56,363 TREEX-INFO: Preparing data set...
 +2013-01-04 13:31:04,559 TREEX-INFO: Filtering...
 +/home/odusek/.local-x86_64/lib/python2.7/site-packages/sklearn/feature_selection/univariate_selection.py:94: RuntimeWarning: divide by zero encountered in divide  f = msb / msw
 +2013-01-04 13:31:49,558 TREEX-INFO: Training...
 +2013-01-05 00:55:21,004 TREEX-INFO: Training done.
 +2013-01-05 00:55:21,374 TREEX-INFO: Evaluation on file: data/dtest.arff.gz
 +2013-01-05 00:56:10,144 TREEX-INFO: Score: 0.961291064956
 +2013-01-05 00:56:10,149 TREEX-INFO: Saving model to file runs/train-plain_pydata/model.pickle.gz
 +2013-01-05 00:57:22,448 TREEX-INFO: Model successfully saved.
 +
 +** Výsledky pro různé parametry logreg **
 +
 +  * Nepomáhá příliš malé C = 0.1, ani příliš malé tol = 0.1
 +  * Spíš ani C = 1 není nic moc, C = 10 nebo 100 je mnohem lepší
 +  * Tol taky radši = 0.001 nebo 0.0001
 +  * Na L2 / L1 druhu regularizace zřejmě moc nezávisí
 +  * Rozpětí 96.92 - 94.01, naprostá většina nad 96.5
 +  * L2 regularizace tvoří nechutně velké modely, L1 jsou mnooohem menší
 +
 +  train-l2_1000_001.py.o6633181:2013-01-11 03:15:37,871 TREEX-INFO: Score: 0.968472611875
 +  train-l1_100_0001.py.o6633154:2013-01-10 18:35:26,878 TREEX-INFO: Score: 0.968484109828
 +  train-l1_100_00001.py.o6633155:2013-01-11 04:25:29,295 TREEX-INFO: Score: 0.968886538196
 +  train-l1_10_0001.py.o6633150:2013-01-11 08:14:22,650 TREEX-INFO: Score: 0.96910499931
 +  train-l1_10_00001.py.o6633151:2013-01-11 03:05:33,517 TREEX-INFO: Score: 0.969254472704
 +
 +** Použití SVM **
 +  * Lineární SVM -- trvá dýl trénování, jinak není rozdíl -- nedosahují ani nejlepších výsledků.
 +    * hlavně s L2 je dlouhé.
 +
 +  train-l1_l2_1_False_0001.py.o6636505:2013-01-12 21:33:26,919 TREEX-INFO: Score: 0.964501467119
 +  train-l2_l2_1_False_0001.py.o6636541:2013-01-14 16:01:52,936 TREEX-INFO: Score: 0.964363385306
 +  train-l2_l2_10_False_0001.py.o6636544:2013-01-14 06:30:18,333 TREEX-INFO: Score: 0.964363385306
 +  train-l1_l2_1_False_00001.py.o6636506:2013-01-13 09:29:26,141 TREEX-INFO: Score: 0.964363385306
  

[ Back to the navigation ] [ Back to the content ]