Ensembles structure fix.

w4k2 · Dec 11, 2019 · a3ccd4b · a3ccd4b
1 parent 7c7111c
commit a3ccd4b
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 4 deletions.
diff --git a/doc/ensembles.rst b/doc/ensembles.rst
@@ -20,9 +20,9 @@ ensembles can be categorized as `chunk-based` or `online`. The ``stream-learn``
 package implements various ensemble methods for data stream classification,
 which can be found in the ``ensembles`` module.
 
-######################################
+
 Chunk-based Ensembles for Data Streams
-######################################
+======================================
 
 Chunk-based approaches process successively incoming data chunks containing a predetermined number of instances. The learning algorithm can repeatedly process training samples located in a given data chunk to learn base models. It is worth noting that this does not mean that batch processing can only be used when new instances arrive in chunks. These approaches can also be used when instances arrive individually, if we store each new sample in a buffer until its size is equal to the size of the chunk.
 
@@ -34,9 +34,8 @@ Weighted Aging Ensemble (WAE)
 -----------------------------
 
 
-#################################
 Online Ensembles for Data Streams
-#################################
+=================================
 
 Online approaches, unlike those based on batch processing, process each new sample separately. These methods have been developed for applications with memory and computational limitations (i.e. where the amount of incoming data is extensive). Online methods can also be used in cases where data samples do not arrive separately. These types of methods can process each instances of data chunk are individually and can therefore be used in an environment where data arrives in batches.
 

diff --git a/examples/disco.png b/examples/disco.png