【机器学习】1.3.k均值聚类

包含全部示例的代码仓库见GIthub

1 导入库

import pandas as pd
import matplotlib.pyplot as plt
%matplotlib inline
import numpy as np
from sklearn.cluster import KMeans
from sklearn import metrics

2 数据准备

data1 = pd.DataFrame({'X':np.random.randint(1,50,100), 'Y':np.random.randint(1,50,100)})
data = pd.concat([data1+50, data1])
data

绘图

plt.style.use('ggplot')
plt.scatter(data.X, data.Y)

3 模型构建

y_pred = KMeans(n_clusters=2).fit_predict(data)
y_pred
# output
array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0])

plt.scatter(data.X, data.Y, c=y_pred)

4 模型评价与改进

评价是否聚集

metrics.calinski_harabasz_score(data, y_pred)  # 评价是否聚集，值越大效果越好
# output
626.3642402960996

y_pred = KMeans(n_clusters=3).fit_predict(data)
plt.scatter(data.X, data.Y, c=y_pred)

metrics.calinski_harabasz_score(data, y_pred)
# output
437.4815970504506

y_pred = KMeans(n_clusters=4).fit_predict(data)
metrics.calinski_harabasz_score(data, y_pred)
# output
447.54048271842913

改进后评价指标下降，说明应n_clusters=2

4 多变量情况

数据准备

# delim_whitespace=True 空格为分隔符
data = pd.read_csv('./dataset/seeds_dataset.txt', 
                   header=None, delim_whitespace=True,
                   names=['x1', 'x2', 'x3', 'x4', 'x5', 'x6', 'x7', 'y'])
data
# output
        x1	x2	x3	x4	x5	x6	x7	y
0	15.26	14.84	0.8710	5.763	3.312	2.221	5.220	1
1	14.88	14.57	0.8811	5.554	3.333	1.018	4.956	1
2	14.29	14.09	0.9050	5.291	3.337	2.699	4.825	1
3	13.84	13.94	0.8955	5.324	3.379	2.259	4.805	1
4	16.14	14.99	0.9034	5.658	3.562	1.355	5.175	1
...	...	...	...	...	...	...	...	...
205	12.19	13.20	0.8783	5.137	2.981	3.631	4.870	3
206	11.23	12.88	0.8511	5.140	2.795	4.325	5.003	3
207	13.20	13.66	0.8883	5.236	3.232	8.315	5.056	3
208	11.84	13.21	0.8521	5.175	2.836	3.598	5.044	3
209	12.30	13.34	0.8684	5.243	2.974	5.637	5.063	3

模型构建

y_pred = KMeans(n_clusters=3).fit_predict(data[['x1', 'x2', 'x3', 'x4', 'x5', 'x6', 'x7']])
y_pre = pd.Series(y_pred)
y_pre = y_pre.map({0:1,1:3})
y_pre
# output
0      1.0
1      1.0
2      1.0
3      1.0
4      1.0
      ... 
205    NaN
206    NaN
207    NaN
208    NaN
209    NaN

data.y
# output
0      1
1      1
2      1
3      1
4      1
      ..
205    3
206    3
207    3
208    3
209    3

评价

data_p = pd.DataFrame({'y_pre':y_pre, 'y':data.y})、
data_p['acc'] = data_p.y_pre==data_p.y
data_p.acc.sum()/len(data_p)
# output
0.2857142857142857

metrics.calinski_harabasz_score(data[['x1', 'x2', 'x3', 'x4', 'x5', 'x6', 'x7']], y_pred)  # 评价
# output
375.8049613895007

y_pred = KMeans(n_clusters=4).fit_predict(data[['x1', 'x2', 'x3', 'x4', 'x5', 'x6', 'x7']])
metrics.calinski_harabasz_score(data[['x1', 'x2', 'x3', 'x4', 'x5', 'x6', 'x7']], y_pred)  # 评价
# output
327.43909895952993

y_pred = KMeans(n_clusters=5).fit_predict(data[['x1', 'x2', 'x3', 'x4', 'x5', 'x6', 'x7']])
metrics.calinski_harabasz_score(data[['x1', 'x2', 'x3', 'x4', 'x5', 'x6', 'x7']], y_pred)  # 评
# output
310.331839794128