机器学习--K-means&DBSCAN

一. K-means实现

1. 读取数据

import pandas as pd
beer = pd.read_csv('data.txt', sep=' ')
beer

2. 取出其中四列

X = beer[["calories","sodium","alcohol","cost"]]
X

3. 使用K-means对X进行分类

from sklearn.cluster import KMeans
# 分别取3个和2个质心的情况
km = KMeans(n_clusters=3).fit(X)
km2 = KMeans(n_clusters=2).fit(X)
km.labels_ # 每个样本点对应的分类

4. 根据分类进行排序

beer['cluster'] = km.labels_
beer['cluster2'] = km2.labels_
beer.sort_values('cluster')

5. 计算cluster=3时的平均值

beer.groupby('cluster').mean()

6. 计算cluster=2时的平均值

beer.groupby('cluster2').mean()

7. 利用reset_index重置索引

centers = beer.groupby('cluster').mean().reset_index()
centers

8. 分类效果的可视化

%matplotlib inline
import numpy as np
import matplotlib.pyplot as plt

plt.rcParams['font.size'] = 14
colors = np.array(['red', 'green', 'blue', 'yellow'])
# 数据分布三点图
plt.scatter(beer['calories'], beer['alcohol'], c=colors[beer['cluster']]) 
# 关于 calories alcohol 的三个类的平均值分布
plt.scatter(centers.calories, centers.alcohol, linewidth=3, marker='+', s=300, c='black') 

plt.xlabel('Calories')
plt.ylabel('Alcohol')

9. 计算不同cluster每个分类不同特征的centers值

from pandas.tools.plotting import scatter_matrix
%matplotlib inline

cluster_centers = km.cluster_centers_
cluster_centers_2 = km2.cluster_centers_

print (cluster_centers)
print (cluster_centers_2)

10. 基于步骤9，cluster=3 进行可视化

scatter_matrix(beer[["calories","sodium","alcohol","cost"]],s=100, alpha=1, c=colors[beer["cluster"]], figsize=(10,10))
plt.suptitle("With 3 centroids initialized")

10. 基于步骤9，cluster=2 进行可视化

scatter_matrix(beer[["calories","sodium","alcohol","cost"]],s=100, alpha=1, c=colors[beer["cluster2"]], figsize=(10,10))
plt.suptitle("With 2 centroids initialized")

二. 数据标准化

1. 数据标准化

from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
X_scaled = scaler.fit_transform(X)
X_scaled

2. 使用标准化后的数据生成新的cluster=3，且应用于原始数据进行分类

km = KMeans(n_clusters=3).fit(X_scaled)
beer["scaled_cluster"] = km.labels_
beer.sort_values("scaled_cluster")

3. 求mean值

beer.groupby("scaled_cluster").mean()

4. 可视化

pd.scatter_matrix(X, c=colors[beer.scaled_cluster], alpha=1, figsize=(10,10), s=100)

三：聚类评估：轮廓系数（Silhouette Coefficient ）

1. 计算当前分类下的轮廓系数

from sklearn import metrics
score_scaled = metrics.silhouette_score(X,beer.scaled_cluster) #数据标准化的轮廓系数
score = metrics.silhouette_score(X,beer.cluster) #原始数据的轮廓系数
print(score_scaled, score)

2. 计算不同cluster值下的轮廓系数

scores = []
for k in range(2,20):
    labels = KMeans(n_clusters=k).fit(X).labels_
    score = metrics.silhouette_score(X, labels)
    scores.append(score)

scores

3. 基于2，进行可视化，观察走向（cluster=2最合适）

plt.plot(list(range(2,20)), scores)
plt.xlabel("Number of Clusters Initialized")
plt.ylabel("Sihouette Score")

四.DBSCAN Clustering

1. 数据初始化，使用DBSCAN进行分类

from sklearn.cluster import DBSCAN
db = DBSCAN(eps=10, min_samples=2).fit(X) #eps 半径 min_samples 最小密度
labels = db.labels_
beer['cluster_db'] = labels
beer.sort_values('cluster_db')

2. 求每个特征的均值

beer.groupby('cluster_db').mean()

3. 可视化

pd.scatter_matrix(X, c=colors[beer.cluster_db], figsize=(10,10), s=100)

4. 数据标准化后进行DBSCAN分类

db = DBSCAN(eps=10, min_samples=2).fit(X) #eps 半径 min_samples 最小密度
labels = db.labels_
beer['cluster_db'] = labels
beer.sort_values('cluster_db')

5. 通过改变eps和min_samples数值，计算轮廓系数并可视化

scores=[]
counter=0
for eps in range(10,20):
    for min_samples in range(2,10):
        labels=DBSCAN(eps,min_samples).fit(X).labels_
        score=metrics.silhouette_score(X,labels)
        scores.append(score)
        counter+=1
# scores
counter
plt.plot((range(0,80)), scores)
plt.xlabel("value of eps Initialized")
plt.ylabel("Sihouette Score")