无监督学习：聚类/降维/异常检测

1. 聚类算法

fromsklearn.clusterimportKMeans,DBSCAN,AgglomerativeClustering# K-Meanskmeans=KMeans(n_clusters=3,random_state=42)labels=kmeans.fit_predict(X)# 肘部法则选择 Kinertias=[]forkinrange(2,11):km=KMeans(n_clusters=k,random_state=42)km.fit(X)inertias.append(km.inertia_)# DBSCAN（密度聚类）dbscan=DBSCAN(eps=0.5,min_samples=5)labels=dbscan.fit_predict(X)# 层次聚类hc=AgglomerativeClustering(n_clusters=3)labels=hc.fit_predict(X)

2. 降维算法

fromsklearn.decompositionimportPCAfromsklearn.manifoldimportTSNE# PCApca=PCA(n_components=2)X_pca=pca.fit_transform(X)print(f"解释方差比:{pca.explained_variance_ratio_}")# t-SNE（可视化用）tsne=TSNE(n_components=2,random_state=42,perplexity=30)X_tsne=tsne.fit_transform(X)

3. 异常检测

fromsklearn.ensembleimportIsolationForestfromsklearn.svmimportOneClassSVM# 孤立森林iso=IsolationForest(contamination=0.1,random_state=42)outliers=iso.fit_predict(X)# -1 为异常# One-Class SVMocsvm=OneClassSVM(kernel='rbf',nu=0.1)outliers=ocsvm.fit_predict(X)

总结