[博客精选]Implementing your own k-nearest neighbour algorithm using Python

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

In machine learning, you may often wish to build predictors that allows to classify things into categories based on some set of associated values. For example, it is possible to provide a diagnosis to a patient based on data from previous patients.

Classification can involve constructing highly non-linear boundaries between classes, as in the case of the red, green and blue classes below:

本帖隐藏的内容

7.pdf (5.2 MB)

2016-8-19 01:33:16 上传

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：Implementing implement Algorithm neighbour python neighbour

1. from sklearn.datasets import load_iris
   
2. from sklearn import cross_validation
   
3. from sklearn.metrics import classification_report, accuracy_score
   
4. from operator import itemgetter
   
5. import numpy as np
   
6. import math
   
7. from collections import Counter
   
8. 
   
9. # 1) given two data points, calculate the euclidean distance between them
   
10. def get_distance(data1, data2):
    
11.     points = zip(data1, data2)
    
12.     diffs_squared_distance = [pow(a - b, 2) for (a, b) in points]
    
13.     return math.sqrt(sum(diffs_squared_distance))
    
14. 
    
15. # 2) given a training set and a test instance, use getDistance to calculate all pairwise distances
    
16. def get_neighbours(training_set, test_instance, k):
    
17.     distances = [_get_tuple_distance(training_instance, test_instance) for training_instance in training_set]
    
18.     # index 1 is the calculated distance between training_instance and test_instance
    
19.     sorted_distances = sorted(distances, key=itemgetter(1))
    
20.     # extract only training instances
    
21.     sorted_training_instances = [tuple[0] for tuple in sorted_distances]
    
22.     # select first k elements
    
23.     return sorted_training_instances[:k]
    
24. 
    
25. def _get_tuple_distance(training_instance, test_instance):
    
26.     return (training_instance, get_distance(test_instance, training_instance[0]))
    
27. 
    
28. # 3) given an array of nearest neighbours for a test case, tally up their classes to vote on test case class
    
29. def get_majority_vote(neighbours):
    
30.     # index 1 is the class
    
31.     classes = [neighbour[1] for neighbour in neighbours]
    
32.     count = Counter(classes)
    
33.     return count.most_common()[0][0]
    
34. 
    
35. # setting up main executable method
    
36. def main():
    
37. 
    
38.     # load the data and create the training and test sets
    
39.     # random_state = 1 is just a seed to permit reproducibility of the train/test split
    
40.     iris = load_iris()
    
41.     X_train, X_test, y_train, y_test = cross_validation.train_test_split(iris.data, iris.target, test_size=0.4, random_state=1)
    
42. 
    
43.     # reformat train/test datasets for convenience
    
44.     train = np.array(zip(X_train,y_train))
    
45.     test = np.array(zip(X_test, y_test))
    
46. 
    
47.     # generate predictions
    
48.     predictions = []
    
49. 
    
50.     # let's arbitrarily set k equal to 5, meaning that to predict the class of new instances,
    
51.     k = 5
    
52. 
    
53.     # for each instance in the test set, get nearest neighbours and majority vote on predicted class
    
54.     for x in range(len(X_test)):
    
55. 
    
56.             print 'Classifying test instance number ' + str(x) + ":",
    
57.             neighbours = get_neighbours(training_set=train, test_instance=test[x][0], k=5)
    
58.             majority_vote = get_majority_vote(neighbours)
    
59.             predictions.append(majority_vote)
    
60.             print 'Predicted label=' + str(majority_vote) + ', Actual label=' + str(test[x][1])
    
61. 
    
62.     # summarize performance of the classification
    
63.     print '\nThe overall accuracy of the model is: ' + str(accuracy_score(y_test, predictions)) + "\n"
    
64.     report = classification_report(y_test, predictions, target_names = iris.target_names)
    
65.     print 'A detailed classification report: \n\n' + report
    
66. 
    
67. if __name__ == "__main__":
    
68.     main()

复制代码

kankan