Pushing the docs to dev/ for branch: main, commit f16c987592d01890cf61679f7f323bc15cbd1269

Author: sklearn-ci

dev/_downloads/07fcc19ba03226cd3d83d4e40ec44385/auto_examples_python.zip

@@ -94,7 +94,7 @@
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "In the figure above, each vertex of the simplex represents\na perfectly predicted class (e.g., 1, 0, 0). The mid point\ninside the simplex represents predicting the three classes with equal\nprobability (i.e., 1/3, 1/3, 1/3). Each arrow starts at the\nuncalibrated probabilities and end with the arrow head at the calibrated\nprobability. The color of the arrow represents the true class of that test\nsample.\n\nThe uncalibrated classifier is overly confident in its predictions and\nincurs a large `log loss <log_loss>`. The calibrated classifier incurs\na lower `log loss <log_loss>` due to two factors. First, notice in the\nfigure above that the arrows generally point away from the edges of the\nsimplex, where the probability of one class is 0. Second, a large proportion\nof the arrows point towards the true class, e.g., green arrows (samples where\nthe true class is 'green') generally point towards the green vertex. This\nresults in fewer over-confident, 0 predicted probabilities and at the same\ntime an increase in the predicted probabilities of the correct class.\nThus, the calibrated classifier produces more accurate predicted probablities\nthat incur a lower `log loss <log_loss>`\n\nWe can show this objectively by comparing the `log loss <log_loss>` of\nthe uncalibrated and calibrated classifiers on the predictions of the 1000\ntest samples. Note that an alternative would have been to increase the number\nof base estimators (trees) of the\n:class:`~sklearn.ensemble.RandomForestClassifier` which would have resulted\nin a similar decrease in `log loss <log_loss>`.\n\n"
+        "In the figure above, each vertex of the simplex represents\na perfectly predicted class (e.g., 1, 0, 0). The mid point\ninside the simplex represents predicting the three classes with equal\nprobability (i.e., 1/3, 1/3, 1/3). Each arrow starts at the\nuncalibrated probabilities and end with the arrow head at the calibrated\nprobability. The color of the arrow represents the true class of that test\nsample.\n\nThe uncalibrated classifier is overly confident in its predictions and\nincurs a large `log loss <log_loss>`. The calibrated classifier incurs\na lower `log loss <log_loss>` due to two factors. First, notice in the\nfigure above that the arrows generally point away from the edges of the\nsimplex, where the probability of one class is 0. Second, a large proportion\nof the arrows point towards the true class, e.g., green arrows (samples where\nthe true class is 'green') generally point towards the green vertex. This\nresults in fewer over-confident, 0 predicted probabilities and at the same\ntime an increase in the predicted probabilities of the correct class.\nThus, the calibrated classifier produces more accurate predicted probabilities\nthat incur a lower `log loss <log_loss>`\n\nWe can show this objectively by comparing the `log loss <log_loss>` of\nthe uncalibrated and calibrated classifiers on the predictions of the 1000\ntest samples. Note that an alternative would have been to increase the number\nof base estimators (trees) of the\n:class:`~sklearn.ensemble.RandomForestClassifier` which would have resulted\nin a similar decrease in `log loss <log_loss>`.\n\n"
       ]
     },
     {

dev/_downloads/6f1e7a639e0699d6164445b55e6c116d/auto_examples_jupyter.zip

@@ -198,7 +198,7 @@ class of an instance (red: class 1, green: class 2, blue: class 3).
 # the true class is 'green') generally point towards the green vertex. This
 # results in fewer over-confident, 0 predicted probabilities and at the same
 # time an increase in the predicted probabilities of the correct class.
-# Thus, the calibrated classifier produces more accurate predicted probablities
+# Thus, the calibrated classifier produces more accurate predicted probabilities
 # that incur a lower :ref:`log loss <log_loss>`
 #
 # We can show this objectively by comparing the :ref:`log loss <log_loss>` of