Clarify definition of cross-entropy metric in the documentation (clea…

…n up PR #7291) (#7365)

* [R] switch order of LRN and pooling layer

Original paper (section 3.5) performs local response normalization of relu.

* clearify definition of cross entropy

* fix small type

* fixed lint, trailing wspace

python/mxnet/metric.py

@@ -854,10 +854,14 @@ def update(self, labels, preds):
 class CrossEntropy(EvalMetric):
     """Computes Cross Entropy loss.
 
-    The cross entropy is given by
+    The cross entropy over a batch of sample size :math:`N` is given by
 
     .. math::
-        -y\\log \\hat{y} + (1-y)\\log (1-\\hat{y})
+       -\\sum_{n=1}^{N}\\sum_{k=1}^{K}t_{nk}\\log (y_{nk}),
+
+    where :math:`t_{nk}=1` if and only if sample :math:`n` belongs to class :math:`k`.
+    :math:`y_{nk}` denotes the probability of sample :math:`n` belonging to
+    class :math:`k`.
 
     Parameters
     ----------