Python k-Nearest Neighbor Algorithm instance sharing

Simple Description

This algorithm is mainly used to measure the distance between different feature values. With this distance, you can classify them.

KNN for short.

Known: the training set and the label of each training set.

Next, compare with the data in the training set to calculate the most similar k Distance. Select the category with the most similarity data. As the classification of new data.

Python instance

Copy codeThe Code is as follows:
#-*-Coding: cp936 -*-

# In Windows, cp936 encoding is used. It is better to use UTF-8 in linux.
From numpy import * # introduce the scientific computing package
Import operator # classic python function library. Operator module.

# Create a dataset
Def createDataSet ():
Group = array ([[1.0, 1.1], [1.0, 1.0], [0.1], [0,])
Labels = ['A', 'A', 'B', 'B']
Return group, labels

# Algorithm core
# Classification: input vector used for classification. It will be classified soon.
# DataSet: training sample set
# Labels: Label Vector
Def classfy0 (datasets, dataSet, labels, k ):
# Distance Calculation
DataSetSize = dataSet. shape [0] # obtain the number of rows in the array. I know that there are several training data sets.
DiffMat = tile (partition, (dataSetSize, 1)-dataSet # tile: Functions in numpy. Tile expands the original array into four identical arrays. DiffMat obtains the difference between the target and the training value.
SqDiffMat = diffMat ** 2 # each element is square
SqDistances = sqDiffMat. sum (axis = 1) # multiply the corresponding column to obtain the square of each distance.
Distances = sqDistances ** 0.5 # Start, get the distance.
SortedDistIndicies = distances. argsort () # sort in ascending order
# Select the nearest k points.
ClassCount = {}
For I in range (k ):
VoteIlabel = labels [sortedDistIndicies [I]
ClassCount [voteIlabel] = classCount. get (voteIlabel, 0) + 1
# Sorting
SortedClassCount = sorted (classCount. iteritems (), key = operator. itemgetter (1), reverse = True)
Return sortedClassCount [0] [0]

Unexpected gains

Add a self-written module to the default search path of python: Create a xxx. pth file under the python/lib/-packages directory and write it to the path of the Self-written module.