Improve docs

lib/scholar/neighbors/k_nearest_neighbors.ex

@@ -3,7 +3,9 @@ defmodule Scholar.Neighbors.KNearestNeighbors do
   The K-Nearest Neighbors.
 
   It implements both classification and regression. This implements the linear
-  version of kNN and therefore it has time complexity $O(N^2)$ for $N$ samples.
+  (brute) version of kNN and therefore it has time complexity $O(N^2)$ for $N$
+  samples. See `Scholar.Neighbors.KDTree` for an alternative implementation which
+  is also exact.
   """
   import Nx.Defn
   import Scholar.Shared

lib/scholar/neighbors/kd_tree.ex

@@ -3,6 +3,8 @@ defmodule Scholar.Neighbors.KDTree do
   Implements a kd-tree, a space-partitioning data structure for organizing points
   in a k-dimensional space.
 
+  It can be used to predict the K-Nearest Neighbors of a given input.
+
   This is implemented as one-dimensional tensor with indices pointed to highest
   dimension of the given tensor. Traversal starts by calling `root/0` and then
   accessing the `left_child/1` and `right_child/1`. The tree is left-balanced.

lib/scholar/neighbors/radius_nearest_neighbors.ex

@@ -163,7 +163,7 @@ defmodule Scholar.Neighbors.RadiusNearestNeighbors do
 
   ## Return Values
 
-    It returns a tensor with predicted class labels
+  It returns a tensor with predicted class labels.
 
   ## Examples
 
@@ -258,7 +258,7 @@ defmodule Scholar.Neighbors.RadiusNearestNeighbors do
 
   ## Return Values
 
-    Returns indices of the selected neighbor points as a mask  (1 if a point is a neighbor, 0 otherwise) and their respective distances.
+  Returns indices of the selected neighbor points as a mask  (1 if a point is a neighbor, 0 otherwise) and their respective distances.
 
   ## Examples