DOC Use defined notation for precision and recall (#12726)

Greg Gandenberger · qinhanmin2014 · commit 1cb56ba882a9 · 2018-12-06T09:10:23.000+08:00
diff --git a/doc/modules/model_evaluation.rst b/doc/modules/model_evaluation.rst
@@ -887,11 +887,11 @@ To make this more explicit, consider the following notation:
 * :math:`y_l` the subset of :math:`y` with label :math:`l`
 * similarly, :math:`\hat{y}_s` and :math:`\hat{y}_l` are subsets of
   :math:`\hat{y}`
-* :math:`P(A, B) := \frac{\left| A \cap B \right|}{\left|A\right|}`
-* :math:`R(A, B) := \frac{\left| A \cap B \right|}{\left|B\right|}`
-  (Conventions vary on handling :math:`B = \emptyset`; this implementation uses
-  :math:`R(A, B):=0`, and similar for :math:`P`.)
-* :math:`F_\beta(A, B) := \left(1 + \beta^2\right) \frac{P(A, B) \times R(A, B)}{\beta^2 P(A, B) + R(A, B)}`
+* :math:`P(y_l, \hat{y}_l) := \frac{\left| y_l \cap \hat{y}_l \right|}{\left|y_l\right|}`
+* :math:`R(y_l, \hat{y}_l) := \frac{\left| y_l \cap \hat{y}_l \right|}{\left|\hat{y}_l\right|}`
+  (Conventions vary on handling :math:`\hat{y}_l = \emptyset`; this implementation uses
+  :math:`R(y_l, \hat{y}_l):=0`, and similar for :math:`P`.)
+* :math:`F_\beta(y_l, \hat{y}_l) := \left(1 + \beta^2\right) \frac{P(y_l, \hat{y}_l) \times R(y_l, \hat{y}_l)}{\beta^2 P(y_l, \hat{y}_l) + R(y_l, \hat{y}_l)}`
 
 Then the metrics are defined as:
 
