这是什么程序？What is this ?

不同于scikit-learn的NaiveBayes输入只能是离散型变量或连续性变量, naive_bayes_plus输入既可以有离散型变量也可以有连续性变量的

Different with NaiveBayes of scikit-learn which just accept discrete attributes or continuous attributes as input, the naive_bayes_plus could accept discrete attributes and continuous attributes as input in same time.

1 中文版(简体)

1.1 运行环境

Python3

1.2 Python依赖包

包名	功能	Pip 安装命令
numpy	最流行的数值计算包	Pip install numpy

1.3 如何运行?

下面是一个训练集

性别	身高(英尺)	体重(磅)	脚掌(英寸)	喜欢玩游戏？
男	6	180	12	否
男	5.92	190	11	是
男	5.58	170	12	是
男	5.92	165	10	是
女	5	100	6	否
女	5.5	150	8	否
女	5.42	130	7	否
女	5.75	150	9	是

请用朴素贝叶斯算法预测下面这个人是男还是女？

身高=6
体重=130
脚掌=8
喜欢玩游戏？=是

输入的Python代码

from naive_bayes_plus import NavieBayesPlus

nbp = NavieBayesPlus()
l_train_x_dat = [] 
l_train_x_dat.append([6, 180, 12, 'False'])
l_train_x_dat.append([5.92, 190, 11, 'True'])
l_train_x_dat.append([5.58, 170, 12, 'True'])
l_train_x_dat.append([5.92, 165, 10, 'Trueb'])
l_train_x_dat.append([5, 100, 16, 'False'])
l_train_x_dat.append([5.5, 150, 8, 'False'])
l_train_x_dat.append([5.42, 130, 7, 'False'])
l_train_x_dat.append([5.75, 150, 9, 'True'])

l_train_y_dat = ['Male','Male','Male','Male','Female','Female','Female','Female']

nbp.train(l_train_x_dat, l_train_y_dat)
l_y, l_y_prob = nbp.predict([[6, 130, 8, 'True']])
print(l_y)
print(l_y_prob)

程序输出为

['Female']
[{'Male': 9.917348375732059e-11, 'Female': 1.1457231805275285e-07}]

更多细节请看_test_code.py

2 Run Enviroment

2.1 Run Enviroment

Python3

1.2 Python package

name	function	Pip command
numpy	most popular numercial calculation package	Pip install numpy

1.3 How to run?

This is a train set

Gender	Hight	Weight	Footer	like playing game?
Male	6	180	12	False
Male	5.92	190	11	True
Male	5.58	170	12	True
Male	5.92	165	10	True
Female	5	100	6	False
Female	5.5	150	8	False
Female	5.42	130	7	False
Female	5.75	150	9	True

So please use Naive-Bayes to predict the person is male or female.

Hight=6
Weight=130
Footer=8
like playing game?=True

Input Python Code

from naive_bayes_plus import NavieBayesPlus

nbp = NavieBayesPlus()
l_train_x_dat = [] 
l_train_x_dat.append([6, 180, 12, 'False'])
l_train_x_dat.append([5.92, 190, 11, 'True'])
l_train_x_dat.append([5.58, 170, 12, 'True'])
l_train_x_dat.append([5.92, 165, 10, 'Trueb'])
l_train_x_dat.append([5, 100, 16, 'False'])
l_train_x_dat.append([5.5, 150, 8, 'False'])
l_train_x_dat.append([5.42, 130, 7, 'False'])
l_train_x_dat.append([5.75, 150, 9, 'True'])

l_train_y_dat = ['Male','Male','Male','Male','Female','Female','Female','Female']

nbp.train(l_train_x_dat, l_train_y_dat)
l_y, l_y_prob = nbp.predict([[6, 130, 8, 'True']])
print(l_y)
print(l_y_prob)

Output

['Female']
[{'Male': 9.917348375732059e-11, 'Female': 1.1457231805275285e-07}]

more details in _test_code.py

Yumeka999/python-naive_bayes_plus