重庆分公司,新征程启航

为企业提供网站建设、域名注册、服务器等服务

R语言朴素贝叶斯技术怎么使用

本篇内容主要讲解“R语言朴素贝叶斯技术怎么使用”,感兴趣的朋友不妨来看看。本文介绍的方法操作简单快捷,实用性强。下面就让小编来带大家学习“R语言朴素贝叶斯技术怎么使用”吧!

网站建设哪家好,找创新互联公司!专注于网页设计、网站建设、微信开发、小程序开发、集团企业网站建设等服务项目。为回馈新老客户创新互联还提供了沙依巴克免费建站欢迎大家使用!

安装package:

> install.packages("e1071")

导入e1071:

> library(e1071)

找一个数据集:

> data(iris)
> iris
    Sepal.Length Sepal.Width Petal.Length Petal.Width    Species
1            5.1         3.5          1.4         0.2     setosa
2            4.9         3.0          1.4         0.2     setosa
3            4.7         3.2          1.3         0.2     setosa
4            4.6         3.1          1.5         0.2     setosa
5            5.0         3.6          1.4         0.2     setosa
6            5.4         3.9          1.7         0.4     setosa
7            4.6         3.4          1.4         0.3     setosa
8            5.0         3.4          1.5         0.2     setosa
9            4.4         2.9          1.4         0.2     setosa
10           4.9         3.1          1.5         0.1     setosa
11           5.4         3.7          1.5         0.2     setosa
12           4.8         3.4          1.6         0.2     setosa
13           4.8         3.0          1.4         0.1     setosa
14           4.3         3.0          1.1         0.1     setosa
15           5.8         4.0          1.2         0.2     setosa
16           5.7         4.4          1.5         0.4     setosa
17           5.4         3.9          1.3         0.4     setosa
18           5.1         3.5          1.4         0.3     setosa
19           5.7         3.8          1.7         0.3     setosa
20           5.1         3.8          1.5         0.3     setosa
21           5.4         3.4          1.7         0.2     setosa
22           5.1         3.7          1.5         0.4     setosa
23           4.6         3.6          1.0         0.2     setosa
24           5.1         3.3          1.7         0.5     setosa
25           4.8         3.4          1.9         0.2     setosa
26           5.0         3.0          1.6         0.2     setosa
27           5.0         3.4          1.6         0.4     setosa
28           5.2         3.5          1.5         0.2     setosa
29           5.2         3.4          1.4         0.2     setosa
30           4.7         3.2          1.6         0.2     setosa
31           4.8         3.1          1.6         0.2     setosa
32           5.4         3.4          1.5         0.4     setosa
33           5.2         4.1          1.5         0.1     setosa
34           5.5         4.2          1.4         0.2     setosa
35           4.9         3.1          1.5         0.2     setosa
36           5.0         3.2          1.2         0.2     setosa
37           5.5         3.5          1.3         0.2     setosa
38           4.9         3.6          1.4         0.1     setosa
39           4.4         3.0          1.3         0.2     setosa
40           5.1         3.4          1.5         0.2     setosa
41           5.0         3.5          1.3         0.3     setosa
42           4.5         2.3          1.3         0.3     setosa
43           4.4         3.2          1.3         0.2     setosa
44           5.0         3.5          1.6         0.6     setosa
45           5.1         3.8          1.9         0.4     setosa
46           4.8         3.0          1.4         0.3     setosa
47           5.1         3.8          1.6         0.2     setosa
48           4.6         3.2          1.4         0.2     setosa
49           5.3         3.7          1.5         0.2     setosa
50           5.0         3.3          1.4         0.2     setosa
51           7.0         3.2          4.7         1.4 versicolor
52           6.4         3.2          4.5         1.5 versicolor
53           6.9         3.1          4.9         1.5 versicolor
54           5.5         2.3          4.0         1.3 versicolor
55           6.5         2.8          4.6         1.5 versicolor
56           5.7         2.8          4.5         1.3 versicolor
57           6.3         3.3          4.7         1.6 versicolor
58           4.9         2.4          3.3         1.0 versicolor
59           6.6         2.9          4.6         1.3 versicolor
60           5.2         2.7          3.9         1.4 versicolor
61           5.0         2.0          3.5         1.0 versicolor
62           5.9         3.0          4.2         1.5 versicolor
63           6.0         2.2          4.0         1.0 versicolor
64           6.1         2.9          4.7         1.4 versicolor
65           5.6         2.9          3.6         1.3 versicolor
66           6.7         3.1          4.4         1.4 versicolor
67           5.6         3.0          4.5         1.5 versicolor
68           5.8         2.7          4.1         1.0 versicolor
69           6.2         2.2          4.5         1.5 versicolor
70           5.6         2.5          3.9         1.1 versicolor
71           5.9         3.2          4.8         1.8 versicolor
72           6.1         2.8          4.0         1.3 versicolor
73           6.3         2.5          4.9         1.5 versicolor
74           6.1         2.8          4.7         1.2 versicolor
75           6.4         2.9          4.3         1.3 versicolor
76           6.6         3.0          4.4         1.4 versicolor
77           6.8         2.8          4.8         1.4 versicolor
78           6.7         3.0          5.0         1.7 versicolor
79           6.0         2.9          4.5         1.5 versicolor
80           5.7         2.6          3.5         1.0 versicolor
81           5.5         2.4          3.8         1.1 versicolor
82           5.5         2.4          3.7         1.0 versicolor
83           5.8         2.7          3.9         1.2 versicolor
84           6.0         2.7          5.1         1.6 versicolor
85           5.4         3.0          4.5         1.5 versicolor
86           6.0         3.4          4.5         1.6 versicolor
87           6.7         3.1          4.7         1.5 versicolor
88           6.3         2.3          4.4         1.3 versicolor
89           5.6         3.0          4.1         1.3 versicolor
90           5.5         2.5          4.0         1.3 versicolor
91           5.5         2.6          4.4         1.2 versicolor
92           6.1         3.0          4.6         1.4 versicolor
93           5.8         2.6          4.0         1.2 versicolor
94           5.0         2.3          3.3         1.0 versicolor
95           5.6         2.7          4.2         1.3 versicolor
96           5.7         3.0          4.2         1.2 versicolor
97           5.7         2.9          4.2         1.3 versicolor
98           6.2         2.9          4.3         1.3 versicolor
99           5.1         2.5          3.0         1.1 versicolor
100          5.7         2.8          4.1         1.3 versicolor
101          6.3         3.3          6.0         2.5  virginica
102          5.8         2.7          5.1         1.9  virginica
103          7.1         3.0          5.9         2.1  virginica
104          6.3         2.9          5.6         1.8  virginica
105          6.5         3.0          5.8         2.2  virginica
106          7.6         3.0          6.6         2.1  virginica
107          4.9         2.5          4.5         1.7  virginica
108          7.3         2.9          6.3         1.8  virginica
109          6.7         2.5          5.8         1.8  virginica
110          7.2         3.6          6.1         2.5  virginica
111          6.5         3.2          5.1         2.0  virginica
112          6.4         2.7          5.3         1.9  virginica
113          6.8         3.0          5.5         2.1  virginica
114          5.7         2.5          5.0         2.0  virginica
115          5.8         2.8          5.1         2.4  virginica
116          6.4         3.2          5.3         2.3  virginica
117          6.5         3.0          5.5         1.8  virginica
118          7.7         3.8          6.7         2.2  virginica
119          7.7         2.6          6.9         2.3  virginica
120          6.0         2.2          5.0         1.5  virginica
121          6.9         3.2          5.7         2.3  virginica
122          5.6         2.8          4.9         2.0  virginica
123          7.7         2.8          6.7         2.0  virginica
124          6.3         2.7          4.9         1.8  virginica
125          6.7         3.3          5.7         2.1  virginica
126          7.2         3.2          6.0         1.8  virginica
127          6.2         2.8          4.8         1.8  virginica
128          6.1         3.0          4.9         1.8  virginica
129          6.4         2.8          5.6         2.1  virginica
130          7.2         3.0          5.8         1.6  virginica
131          7.4         2.8          6.1         1.9  virginica
132          7.9         3.8          6.4         2.0  virginica
133          6.4         2.8          5.6         2.2  virginica
134          6.3         2.8          5.1         1.5  virginica
135          6.1         2.6          5.6         1.4  virginica
136          7.7         3.0          6.1         2.3  virginica
137          6.3         3.4          5.6         2.4  virginica
138          6.4         3.1          5.5         1.8  virginica
139          6.0         3.0          4.8         1.8  virginica
140          6.9         3.1          5.4         2.1  virginica
141          6.7         3.1          5.6         2.4  virginica
142          6.9         3.1          5.1         2.3  virginica
143          5.8         2.7          5.1         1.9  virginica
144          6.8         3.2          5.9         2.3  virginica
145          6.7         3.3          5.7         2.5  virginica
146          6.7         3.0          5.2         2.3  virginica
147          6.3         2.5          5.0         1.9  virginica
148          6.5         3.0          5.2         2.0  virginica
149          6.2         3.4          5.4         2.3  virginica
150          5.9         3.0          5.1         1.8  virginica

Sepal意思是“花萼 ”,Petal意思是“ 花瓣”。很明显,前四列是花萼和花瓣的特征,第五列代表相应的分类。我们可以用这个数据集进行贝叶斯训练。

先看一下,对这个数据集summary的结果:

> summary(iris)
  Sepal.Length    Sepal.Width     Petal.Length    Petal.Width          Species  
 Min.   :4.300   Min.   :2.000   Min.   :1.000   Min.   :0.100   setosa    :50  
 1st Qu.:5.100   1st Qu.:2.800   1st Qu.:1.600   1st Qu.:0.300   versicolor:50  
 Median :5.800   Median :3.000   Median :4.350   Median :1.300   virginica :50  
 Mean   :5.843   Mean   :3.057   Mean   :3.758   Mean   :1.199                  
 3rd Qu.:6.400   3rd Qu.:3.300   3rd Qu.:5.100   3rd Qu.:1.800                  
 Max.   :7.900   Max.   :4.400   Max.   :6.900   Max.   :2.500

训练并查看训练结果:

> classifier<-naiveBayes(iris[,1:4], iris[,5]) 
> classifier

Naive Bayes Classifier for Discrete Predictors

Call:
naiveBayes.default(x = iris[, 1:4], y = iris[, 5])

A-priori probabilities:
iris[, 5]
    setosa versicolor  virginica 
 0.3333333  0.3333333  0.3333333 

Conditional probabilities:
            Sepal.Length
iris[, 5]     [,1]      [,2]
  setosa     5.006 0.3524897
  versicolor 5.936 0.5161711
  virginica  6.588 0.6358796

            Sepal.Width
iris[, 5]     [,1]      [,2]
  setosa     3.428 0.3790644
  versicolor 2.770 0.3137983
  virginica  2.974 0.3224966

            Petal.Length
iris[, 5]     [,1]      [,2]
  setosa     1.462 0.1736640
  versicolor 4.260 0.4699110
  virginica  5.552 0.5518947

            Petal.Width
iris[, 5]     [,1]      [,2]
  setosa     0.246 0.1053856
  versicolor 1.326 0.1977527
  virginica  2.026 0.2746501

> classifier$apriori
iris[, 5]
    setosa versicolor  virginica 
        50         50         50 
> classifier$tables
$Sepal.Length
            Sepal.Length
iris[, 5]     [,1]      [,2]
  setosa     5.006 0.3524897
  versicolor 5.936 0.5161711
  virginica  6.588 0.6358796

$Sepal.Width
            Sepal.Width
iris[, 5]     [,1]      [,2]
  setosa     3.428 0.3790644
  versicolor 2.770 0.3137983
  virginica  2.974 0.3224966

$Petal.Length
            Petal.Length
iris[, 5]     [,1]      [,2]
  setosa     1.462 0.1736640
  versicolor 4.260 0.4699110
  virginica  5.552 0.5518947

$Petal.Width
            Petal.Width
iris[, 5]     [,1]      [,2]
  setosa     0.246 0.1053856
  versicolor 1.326 0.1977527
  virginica  2.026 0.2746501

classifier中:

A-priori probabilities:
iris[, 5]
    setosa versicolor  virginica 
 0.3333333  0.3333333  0.3333333

很好理解,就是类别的先验概率。
而:

$Petal.Width
            Petal.Width
iris[, 5]     [,1]      [,2]
  setosa     0.246 0.1053856
  versicolor 1.326 0.1977527
  virginica  2.026 0.2746501

是特征Petal.Width的条件概率,在这个贝叶斯实现中,特征是数值型数据(而且还还有小数部分),这里假设概率密度符合高斯分布。比如对于特征Petal.Width,其属于setosa的概率符合mean为0.246,标准方差为0.1053856的高斯分布。



预测:
预测iris数据集中的第一个数据:

> predict(classifier, iris[1, -5])
[1] setosa
Levels: setosa versicolor virginica

iris[1,-5]表示第一行的前4列。

看一下该分类器的效果:

> table(predict(classifier, iris[,-5]), iris[,5], dnn=list('predicted','actual'))
            actual
predicted    setosa versicolor virginica
  setosa         50          0         0
  versicolor      0         47         3
  virginica       0          3        47

分类效果还是不错的。

自己构造一个新的数据并预测:

> new_data = data.frame(Sepal.Length=7, Sepal.Width=3, Petal.Length=6, Petal.Width=2)
> predict(classifier, new_data)
[1] virginica
Levels: setosa versicolor virginica

如果少一个特征(只有三个特征):

> new_data = data.frame(Sepal.Length=7, Sepal.Width=3, Petal.Length=6)
> predict(classifier, new_data)
[1] virginica
Levels: setosa versicolor virginica

下面看一下,这个库如何处理标称型特征:

数据如下:

> model = c("H", "H", "H", "H", "T", "T", "T", "T")
> place = c("B", "B", "N", "N", "B", "B", "N", "N")
> repairs = c("Y", "N", "Y", "N", "Y", "N", "Y", "N")
> dataset = data.frame(model, place, repairs)
> dataset
  model place repairs
1     H     B       Y
2     H     B       N
3     H     N       Y
4     H     N       N
5     T     B       Y
6     T     B       N
7     T     N       Y
8     T     N       N

贝叶斯之:

> classifier<-naiveBayes(dataset[,1:2], dataset[,3]) 
> classifier

Naive Bayes Classifier for Discrete Predictors

Call:
naiveBayes.default(x = dataset[, 1:2], y = dataset[, 3])

A-priori probabilities:
dataset[, 3]
  N   Y 
0.5 0.5 

Conditional probabilities:
            model
dataset[, 3]   H   T
           N 0.5 0.5
           Y 0.5 0.5

            place
dataset[, 3]   B   N
           N 0.5 0.5
           Y 0.5 0.5

好了,预测一下:

> new_data = data.frame(model="H", place="B")
> predict(classifier, new_data)
[1] N
Levels: N Y

perfect!

补充一下,如果某个数据缺少某些特征:

可以用NA代替该特征:

> model = c("H", "H", "H", "H", "T", "T", "T", "T")
> place = c("B", "B", "N", "N", "B", "B", NA, NA)
> repairs = c("Y", "N", "Y", "N", "Y", "N", "Y", "N")
> dataset = data.frame(model, place, repairs)
> dataset
  model place repairs
1     H     B       Y
2     H     B       N
3     H     N       Y
4     H     N       N
5     T     B       Y
6     T     B       N
7     T         Y
8     T         N

> classifier<-naiveBayes(dataset[,1:2], dataset[,3]) 
> classifier

Naive Bayes Classifier for Discrete Predictors

Call:
naiveBayes.default(x = dataset[, 1:2], y = dataset[, 3])

A-priori probabilities:
dataset[, 3]
  N   Y 
0.5 0.5 

Conditional probabilities:
            model
dataset[, 3]   H   T
           N 0.5 0.5
           Y 0.5 0.5

            place
dataset[, 3]         B         N
           N 0.6666667 0.3333333
           Y 0.6666667 0.3333333

到此,相信大家对“R语言朴素贝叶斯技术怎么使用”有了更深的了解,不妨来实际操作一番吧!这里是创新互联网站,更多相关内容可以进入相关频道进行查询,关注我们,继续学习!


当前名称:R语言朴素贝叶斯技术怎么使用
标题链接:http://cqcxhl.com/article/gjcgpo.html

其他资讯

在线咨询
服务热线
服务热线:028-86922220
TOP