Description
Implement k nearest neighbors (kNN) from scratch in Python. You may use any of the standard packages in the Anaconda distribution (e.g., numpy). Your kNN should use the p-norms to measure distance.
The public interface consists of three methods.
-
The constructor. This should take as optional arguments
k, the number of neighbors to use, with the default value 5, and
p, specifying which p-norm to use. The value p = 1 should be an acceptable value.
-
The fit() method. This builds the classi er from the training data. The training data are numpy arrays. The features are organized with one instance per row.
-
The predict() method. This makes predictions. The input is a numpy array of features, one instance per row. This method should return a numpy array containing the predicted class labels.
You are free to implement any internal methods that you wish.
|
c l a s s kNN : |
|||||||||||||||||||||||||||||
|
def |
i n i t |
( s e l f , |
k=5, |
p =2): |
|||||||||||||||||||||||||
|
’ ’ ’ |
|||||||||||||||||||||||||||||
|
k : t h e number |
o f n e i g h b o r s |
used i n |
c l a s s i f i c a t i o n |
||||||||||||||||||||||||||
|
p : s p e c i f i e s |
t h e p |
norm |
t o |
u s e |
|||||||||||||||||||||||||
|
’ ’ ’ |
|||||||||||||||||||||||||||||
|
pass |
|||||||||||||||||||||||||||||
|
def |
f i t ( s e l f |
, X |
t r a i n , |
y |
t r a i n ) : |
||||||||||||||||||||||||
|
’ ’ ’ |
|||||||||||||||||||||||||||||
|
X |
t r a i n : numpy |
a r r a y |
o f |
f e a t u r e s , |
one i n s t a n c e p e r row |
||||||||||||||||||||||||
|
y |
t r a i n : |
numpy |
a r r a y |
o f |
l a b e l s |
||||||||||||||||||||||||
|
’ ’ ’ |
|||||||||||||||||||||||||||||
|
pass |
|||||||||||||||||||||||||||||
|
def |
p r e d i c t ( s e l f |
, |
X |
t e s t ) : |
|||||||||||||||||||||||||
’ ’ ’
X: numpy a r r a y o f f e a t u r e s , one i n s t a n c e p e r row
’ ’ ’
pass
1
