机器学习 on Clarity

聚类算法分析

Sat, 11 Apr 2026 19:27:54 +0000

1import pandas as pd

1from sklearn.datasets import make_blobs  #足球数据

1                   # 数据量      列数           类别        标准差         随机种子
2X, y = make_blobs(n_samples=150, n_features=4, centers=3)

1data = pd.DataFrame(data=X)
2data[4] =y

1data

        0         1          2         3  4

0 -8.317893 0.698439 -10.349959 4.291286 0 1 -5.663915 7.060548 0.330762 4.576624 1 2 -9.276754 1.625037 -9.837227 3.494463 0 3 -4.981855 2.015353 -9.270422 4.598429 0 4 -3.461616 8.064244 4.563894 1.690651 1 .. ... ... ... ... .. 145 -5.561548 6.836946 1.549655 3.830984 1 146 -6.841547 -2.699957 -8.790234 -0.604204 2 147 -5.903813 7.357668 2.510227 3.125459 1 148 -8.607808 0.461901 -9.125664 4.522260 0 149 -8.503622 2.326839 -8.911981 3.532953 0

矩阵特征向量与奇异值分解

Sat, 11 Apr 2026 19:27:54 +0000

特征向量-直方图

1from scipy.linalg import eig  # 矩阵分解
2import numpy as np   # 生成矩阵
3import matplotlib.pyplot as plt  # 绘图

1A = [[1, 2],#生成一个2*2的矩阵
2     [2, 1]] 
3A

[[1, 2], [2, 1]]

1
2# 特征分解
3evals, evecs = eig(A) #求A的特征值（evals）和特征向量(evecs)

[[ 0.70710678 -0.70710678] [ 0.70710678 0.70710678]]

1evecs = evecs[:, 0], evecs[:, 1]
2print(evecs)

(array([0.70710678, 0.70710678]), array([-0.70710678, 0.70710678]))

 1fig, ax = plt.subplots()
 2for spine in ['left', 'bottom']:#让在左下角的坐标轴经过原点
 3    ax.spines[spine].set_position('zero')
 4    ax.grid(alpha=0.4)
 5xmin, xmax = -3, 3
 6ymin, ymax = -3, 3
 7ax.set(xlim=(xmin, xmax), ylim=(ymin, ymax))
 8for v in evecs:
 9    ax.annotate(s="", xy=v, xytext=(0, 0),
10                arrowprops=dict(facecolor='blue',
11                shrink=0,
12                alpha=0.6,
13                width=0.5))
14x = np.linspace(xmin, xmax, 3)#在指定的间隔内返回均匀间隔的数字
15for v in evecs:
16    a = v[1] / v[0] #沿特征向量方向的单位向量
17    ax.plot(x, a * x, 'r-', lw=0.4)# 参数 lw 表示图线的粗细
18plt.show()

奇异值分解

1import numpy as np
2import matplotlib.pyplot as plt

1words = ["books","dad","stock","value","singular","estate","decomposition"]

1X=np.array([[0,2,1,0,0,0,0],[2,0,0,1,0,1,0],[1,0,0,0,0,0,1],[0,0,1,0,0,0,0],[0,1,0,0,0,0,0],[0,0,0,1,1,0,1],[0,1,0,0,1,0,0],[0,0,0,0,1,1,1]])

1X

array([[0, 2, 1, 0, 0, 0, 0], [2, 0, 0, 1, 0, 1, 0], [1, 0, 0, 0, 0, 0, 1], [0, 0, 1, 0, 0, 0, 0], [0, 1, 0, 0, 0, 0, 0], [0, 0, 0, 1, 1, 0, 1], [0, 1, 0, 0, 1, 0, 0], [0, 0, 0, 0, 1, 1, 1]])

数据挖掘通讲

Sat, 11 Apr 2026 19:27:54 +0000

1import pandas as pd
2import numpy as np
3import warnings
4warnings.filterwarnings('ignore')

1.读取除表头外的前五行\读取数据前五行

1df = pd.read_csv('./titanic_trains.csv',encoding='utf8')
2df.head(5)
3df.shape

(891, 12)

2.缺失值处理(没有要求就遵循80%原则)

2.1.查看缺失值情况，输出缺失值比例、列名、缺失值数量

1for x in df.columns:
2    zb = df[x].isnull().sum()/df.shape[0]*100
3    sl = df[x].isnull().sum()
4    print('缺失值占比为:',round(zb,2),"%,列名:",x,',缺失值数量:',sl)

缺失值占比为: 0.0 %,列名: PassengerId ,缺失值数量: 0 缺失值占比为: 0.0 %,列名: Survived ,缺失值数量: 0 缺失值占比为: 0.0 %,列名: Pclass ,缺失值数量: 0 缺失值占比为: 0.0 %,列名: Name ,缺失值数量: 0 缺失值占比为: 0.0 %,列名: Sex ,缺失值数量: 0 缺失值占比为: 19.87 %,列名: Age ,缺失值数量: 177 缺失值占比为: 0.0 %,列名: SibSp ,缺失值数量: 0 缺失值占比为: 0.0 %,列名: Parch ,缺失值数量: 0 缺失值占比为: 0.0 %,列名: Ticket ,缺失值数量: 0 缺失值占比为: 0.0 %,列名: Fare ,缺失值数量: 0 缺失值占比为: 77.1 %,列名: Cabin ,缺失值数量: 687 缺失值占比为: 0.22 %,列名: Embarked ,缺失值数量: 2

数据挖掘通讲：分类算法

Sat, 11 Apr 2026 19:27:54 +0000

1import pandas as pd
2import numpy as np
3import warnings
4warnings.filterwarnings('ignore')

1.读取除表头外的前五行\读取数据前五行

1df = pd.read_csv('./titanic_trains.csv',encoding='utf8')
2df.head(5)
3df.shape

(891, 12)

2.缺失值处理(没有要求就遵循80%原则)

2.1.查看缺失值情况，输出缺失值比例、列名、缺失值数量

1for x in df.columns:
2    zb = df[x].isnull().sum()/df.shape[0]*100
3    sl = df[x].isnull().sum()
4    print('缺失值占比为:',round(zb,2),"%,列名:",x,',缺失值数量:',sl)

数据挖掘通讲：聚类算法

Sat, 11 Apr 2026 19:27:54 +0000

1注意点：
2   1.birch、SpectralClustering这两个算法非常耗资源可能考试环境中的资源不足，运行时会内存溢出，找监考老师。和她说不运行只截图是否可以。

1.读取数据

1import pandas as pd
2import numpy as np

1df = pd.read_csv('./data.csv',encoding='gbk')
2df.head()

客户编号  套餐品牌  信用等级  是否使用4GUSIM卡  是否4G资费   网龄   当月ARPU  当月MOU    当月DOU  视频流量  \

0 10942 2 5 0 1 204 2201.08 2611 54557 22
1 13382 2 5 0 0 201 2181.71 3371 35250 15
2 4192 2 5 1 1 167 2055.60 6913 5884426 1
3 10908 2 5 1 0 171 1827.33 2157 178070 4
4 14130 2 5 0 1 216 1736.40 4218 358592 35