Merge pull request #215 from AndreaPi/patch-2

Update theoretical_description_classification.rst

Author: vtaquet

doc/theoretical_description_classification.rst

@@ -67,7 +67,7 @@ The conformity scores are computed by summing the ranked scores of each label, f
 label of the observation :
 
 .. math:: 
-   s_i(X_i, Y_i) = \sum^k_{j=1} \hat{\mu}(X_i)_{\pi_j} \quad \text{where} \quad Y_i = \pi_j 
+   s_i(X_i, Y_i) = \sum^k_{j=1} \hat{\mu}(X_i)_{\pi_j} \quad \text{where} \quad Y_i = \pi_k 
 
 
 The quantile :math:`\hat{q}` is then computed the same way as the LABEL method.
@@ -198,4 +198,4 @@ NeurIPS 202 (spotlight), 2020.
 
 [3] Anastasios Nikolas Angelopoulos, Stephen Bates, Michael Jordan and Jitendra Malik.
 "Uncertainty Sets for Image Classifiers using Conformal Prediction."
-International Conference on Learning Representations 2021.
+International Conference on Learning Representations 2021.