2. 無監督學習 · sklearn中文文檔

# 2. 無監督學習 - [2.1. 高斯混合模型](modules/mixture.html) - [2.1.1. 高斯混合](modules/mixture.html#id2) - [2.1.1.1. 優缺點 `GaussianMixture`](modules/mixture.html#gaussianmixture) - [2.1.1.1.1. 優點](modules/mixture.html#id3) - [2.1.1.1.2. 缺點](modules/mixture.html#id4) - [2.1.1.2. 選擇經典高斯混合模型中分量的個數](modules/mixture.html#id5) - [2.1.1.3. 估計算法期望最大化（EM）](modules/mixture.html#em) - [2.1.2. 變分貝葉斯高斯混合](modules/mixture.html#bgmm) - [2.1.2.1. 估計算法: 變分推斷（variational inference）](modules/mixture.html#variational-inference) - [2.1.2.1.1. 優點](modules/mixture.html#id8) - [2.1.2.1.2. 缺點](modules/mixture.html#id9) - [2.1.2.2. The Dirichlet Process（狄利克雷過程）](modules/mixture.html#the-dirichlet-process) - [2.2. 流形學習](modules/manifold.html) - [2.2.1. 介紹](modules/manifold.html#id2) - [2.2.2. Isomap](modules/manifold.html#isomap) - [2.2.2.1. 復雜度](modules/manifold.html#id4) - [2.2.3. 局部線性嵌入](modules/manifold.html#locally-linear-embedding) - [2.2.3.1. 復雜度](modules/manifold.html#id6) - [2.2.4. 改進型局部線性嵌入（MLLE）](modules/manifold.html#mlle) - [2.2.4.1. 復雜度](modules/manifold.html#id7) - [2.2.5. 黑塞特征映射（HE）](modules/manifold.html#he) - [2.2.5.1. 復雜度](modules/manifold.html#id8) - [2.2.6. 譜嵌入](modules/manifold.html#spectral-embedding) - [2.2.6.1. 復雜度](modules/manifold.html#id10) - [2.2.7. 局部切空間對齊（LTSA）](modules/manifold.html#ltsa) - [2.2.7.1. 復雜度](modules/manifold.html#id11) - [2.2.8. 多維尺度分析（MDS）](modules/manifold.html#mds) - [2.2.8.1. 度量 MDS](modules/manifold.html#id13) - [2.2.8.2. 非度量 MDS](modules/manifold.html#id14) - [2.2.9. t 分布隨機鄰域嵌入（t-SNE）](modules/manifold.html#t-t-sne) - [2.2.9.1. 優化 t-SNE](modules/manifold.html#id15) - [2.2.9.2. Barnes-Hut t-SNE](modules/manifold.html#barnes-hut-t-sne) - [2.2.10. 實用技巧](modules/manifold.html#id17) - [2.3. 聚類](modules/clustering.html) - [2.3.1. 聚類方法概述](modules/clustering.html#id2) - [2.3.2. K-means](modules/clustering.html#k-means) - [2.3.2.1. 小批量 K-Means](modules/clustering.html#mini-batch-kmeans) - [2.3.3. Affinity Propagation](modules/clustering.html#affinity-propagation) - [2.3.4. Mean Shift](modules/clustering.html#mean-shift) - [2.3.5. Spectral clustering](modules/clustering.html#spectral-clustering) - [2.3.5.1. 不同的標記分配策略](modules/clustering.html#id10) - [2.3.6. 層次聚類](modules/clustering.html#hierarchical-clustering) - [2.3.6.1. Different linkage type: Ward, complete and average linkage](modules/clustering.html#different-linkage-type-ward-complete-and-average-linkage) - [2.3.6.2. 添加連接約束](modules/clustering.html#id12) - [2.3.6.3. Varying the metric](modules/clustering.html#varying-the-metric) - [2.3.7. DBSCAN](modules/clustering.html#dbscan) - [2.3.8. Birch](modules/clustering.html#birch) - [2.3.9. 聚類性能度量](modules/clustering.html#clustering-evaluation) - [2.3.9.1. 調整后的 Rand 指數](modules/clustering.html#rand) - [2.3.9.1.1. 優點](modules/clustering.html#id24) - [2.3.9.1.2. 缺點](modules/clustering.html#id25) - [2.3.9.1.3. 數學表達](modules/clustering.html#id26) - [2.3.9.2. 基于 Mutual Information （互信息）的分數](modules/clustering.html#mutual-information) - [2.3.9.2.1. 優點](modules/clustering.html#id27) - [2.3.9.2.2. 缺點](modules/clustering.html#id28) - [2.3.9.2.3. 數學公式](modules/clustering.html#id29) - [2.3.9.3. 同質性，完整性和 V-measure](modules/clustering.html#v-measure) - [2.3.9.3.1. 優點](modules/clustering.html#id32) - [2.3.9.3.2. 缺點](modules/clustering.html#id33) - [2.3.9.3.3. 數學表達](modules/clustering.html#id34) - [2.3.9.4. Fowlkes-Mallows 分數](modules/clustering.html#fowlkes-mallows) - [2.3.9.4.1. 優點](modules/clustering.html#id35) - [2.3.9.4.2. 缺點](modules/clustering.html#id36) - [2.3.9.5. Silhouette 系數](modules/clustering.html#silhouette) - [2.3.9.5.1. 優點](modules/clustering.html#id37) - [2.3.9.5.2. 缺點](modules/clustering.html#id38) - [2.3.9.6. Calinski-Harabaz 指數](modules/clustering.html#calinski-harabaz) - [2.3.9.6.1. 優點](modules/clustering.html#id39) - [2.3.9.6.2. 缺點](modules/clustering.html#id40) - [2.4. 雙聚類](modules/biclustering.html) - [2.4.1. Spectral Co-Clustering](modules/biclustering.html#spectral-co-clustering) - [2.4.1.1. 數學公式](modules/biclustering.html#id2) - [2.4.2. Spectral Biclustering](modules/biclustering.html#spectral-biclustering) - [2.4.2.1. 數學表示](modules/biclustering.html#id4) - [2.4.3. Biclustering 評測](modules/biclustering.html#biclustering-evaluation) - [2.5. 分解成分中的信號（矩陣分解問題）](modules/decomposition.html) - [2.5.1. 主成分分析（PCA）](modules/decomposition.html#pca) - [2.5.1.1. 準確的PCA和概率解釋（Exact PCA and probabilistic interpretation）](modules/decomposition.html#pca-exact-pca-and-probabilistic-interpretation) - [2.5.1.2. 增量PCA (Incremental PCA)](modules/decomposition.html#pca-incremental-pca) - [2.5.1.3. PCA 使用隨機SVD](modules/decomposition.html#pca-svd) - [2.5.1.4. 核 PCA](modules/decomposition.html#kernel-pca) - [2.5.1.5. 稀疏主成分分析 ( SparsePCA 和 MiniBatchSparsePCA )](modules/decomposition.html#sparsepca-minibatchsparsepca) - [2.5.2. 截斷奇異值分解和隱語義分析](modules/decomposition.html#lsa) - [2.5.3. 詞典學習](modules/decomposition.html#dictionarylearning) - [2.5.3.1. 帶有預計算詞典的稀疏編碼](modules/decomposition.html#sparsecoder) - [2.5.3.2. 通用詞典學習](modules/decomposition.html#id9) - [2.5.3.3. 小批量字典學習](modules/decomposition.html#minibatchdictionarylearning) - [2.5.4. 因子分析](modules/decomposition.html#fa) - [2.5.5. 獨立成分分析（ICA）](modules/decomposition.html#ica) - [2.5.6. 非負矩陣分解(NMF 或 NNMF)](modules/decomposition.html#nmf-nnmf) - [2.5.6.1. NMF 與 Frobenius 范數](modules/decomposition.html#nmf-frobenius) - [2.5.6.2. 具有 beta-divergence 的 NMF](modules/decomposition.html#beta-divergence-nmf) - [2.5.7. 隱 Dirichlet 分配（LDA）](modules/decomposition.html#dirichlet-lda) - [2.6. 協方差估計](modules/covariance.html) - [2.7. 經驗協方差](modules/covariance.html#id2) - [2.8. 收斂協方差](modules/covariance.html#shrunk-covariance) - [2.8.1. 基本收斂](modules/covariance.html#id4) - [2.8.2. Ledoit-Wolf 收斂](modules/covariance.html#ledoit-wolf) - [2.8.3. Oracle 近似收縮](modules/covariance.html#oracle) - [2.9. 稀疏逆協方差](modules/covariance.html#sparse-inverse-covariance) - [2.10. Robust 協方差估計](modules/covariance.html#robust) - [2.10.1. 最小協方差決定](modules/covariance.html#id11) - [2.11. 新奇和異常值檢測](modules/outlier_detection.html) - [2.11.1. Novelty Detection（新奇檢測）](modules/outlier_detection.html#novelty-detection) - [2.11.2. Outlier Detection（異常值檢測）](modules/outlier_detection.html#id2) - [2.11.2.1. Fitting an elliptic envelope（橢圓模型擬合）](modules/outlier_detection.html#fitting-an-elliptic-envelope) - [2.11.2.2. Isolation Forest（隔離森林）](modules/outlier_detection.html#isolation-forest) - [2.11.2.3. Local Outlier Factor（局部異常系數）](modules/outlier_detection.html#local-outlier-factor) - [2.11.2.4. 一類支持向量機與橢圓模型與隔離森林與局部異常系數](modules/outlier_detection.html#id4) - [2.12. 密度估計](modules/density.html) - [2.12.1. 密度估計: 直方圖](modules/density.html#id2) - [2.12.2. 核密度估計](modules/density.html#kernel-density) - [2.13. 神經網絡模型（無監督）](modules/neural_networks_unsupervised.html) - [2.13.1. 限制波爾茲曼機](modules/neural_networks_unsupervised.html#rbm) - [2.13.1.1. 圖形模型和參數化](modules/neural_networks_unsupervised.html#id3) - [2.13.1.2. 伯努利限制玻爾茲曼機](modules/neural_networks_unsupervised.html#id4) - [2.13.1.3. 隨機最大似然學習](modules/neural_networks_unsupervised.html#sml)