# 3. 模型選擇和評估 - [3.1. 交叉驗證：評估估算器的表現](modules/cross_validation.html) - [3.1.1. 計算交叉驗證的指標](modules/cross_validation.html#id2) - [3.1.1.1. cross\_validate 函數和多度量評估](modules/cross_validation.html#cross-validate) - [3.1.1.2. 通過交叉驗證獲取預測](modules/cross_validation.html#id3) - [3.1.2. 交叉驗證迭代器](modules/cross_validation.html#id4) - [3.1.3. 交叉驗證迭代器–循環遍歷數據](modules/cross_validation.html#iid-cv) - [3.1.3.1. K 折](modules/cross_validation.html#k) - [3.1.3.2. 重復 K-折交叉驗證](modules/cross_validation.html#id6) - [3.1.3.3. 留一交叉驗證 (LOO)](modules/cross_validation.html#loo) - [3.1.3.4. 留 P 交叉驗證 (LPO)](modules/cross_validation.html#p-lpo) - [3.1.3.5. 隨機排列交叉驗證 a.k.a. Shuffle & Split](modules/cross_validation.html#a-k-a-shuffle-split) - [3.1.4. 基于類標簽、具有分層的交叉驗證迭代器](modules/cross_validation.html#id7) - [3.1.4.1. 分層 k 折](modules/cross_validation.html#id8) - [3.1.4.2. 分層隨機 Split](modules/cross_validation.html#split) - [3.1.5. 用于分組數據的交叉驗證迭代器](modules/cross_validation.html#group-cv) - [3.1.5.1. 組 k-fold](modules/cross_validation.html#k-fold) - [3.1.5.2. 留一組交叉驗證](modules/cross_validation.html#id10) - [3.1.5.3. 留 P 組交叉驗證](modules/cross_validation.html#p) - [3.1.5.4. Group Shuffle Split](modules/cross_validation.html#group-shuffle-split) - [3.1.6. 預定義的折疊 / 驗證集](modules/cross_validation.html#id11) - [3.1.7. 交叉驗證在時間序列數據中應用](modules/cross_validation.html#timeseries-cv) - [3.1.7.1. 時間序列分割](modules/cross_validation.html#id13) - [3.1.8. A note on shuffling](modules/cross_validation.html#a-note-on-shuffling) - [3.1.9. 交叉驗證和模型選擇](modules/cross_validation.html#id14) - [3.2. 調整估計器的超參數](modules/grid_search.html) - [3.2.1. 網格追蹤法–窮盡的網格搜索](modules/grid_search.html#id3) - [3.2.2. 隨機參數優化](modules/grid_search.html#randomized-parameter-search) - [3.2.3. 參數搜索技巧](modules/grid_search.html#grid-search-tips) - [3.2.3.1. 指定目標度量](modules/grid_search.html#gridsearch-scoring) - [3.2.3.2. 為評估指定多個指標](modules/grid_search.html#multimetric-grid-search) - [3.2.3.3. 復合估計和參數空間](modules/grid_search.html#id16) - [3.2.3.4. 模型選擇：開發和評估](modules/grid_search.html#id18) - [3.2.3.5. 并行機制](modules/grid_search.html#id19) - [3.2.3.6. 對故障的魯棒性](modules/grid_search.html#id20) - [3.2.4. 暴力參數搜索的替代方案](modules/grid_search.html#alternative-cv) - [3.2.4.1. 模型特定交叉驗證](modules/grid_search.html#id22) - [3.2.4.1.1. `sklearn.linear_model`.ElasticNetCV](modules/generated/sklearn.linear_model.ElasticNetCV.html) - [3.2.4.1.2. `sklearn.linear_model`.LarsCV](modules/generated/sklearn.linear_model.LarsCV.html) - [3.2.4.1.3. `sklearn.linear_model`.LassoCV](modules/generated/sklearn.linear_model.LassoCV.html) - [3.2.4.1.3.1. Examples using `sklearn.linear_model.LassoCV`](modules/generated/sklearn.linear_model.LassoCV.html#examples-using-sklearn-linear-model-lassocv) - [3.2.4.1.4. `sklearn.linear_model`.LassoLarsCV](modules/generated/sklearn.linear_model.LassoLarsCV.html) - [3.2.4.1.4.1. Examples using `sklearn.linear_model.LassoLarsCV`](modules/generated/sklearn.linear_model.LassoLarsCV.html#examples-using-sklearn-linear-model-lassolarscv) - [3.2.4.1.5. `sklearn.linear_model`.LogisticRegressionCV](modules/generated/sklearn.linear_model.LogisticRegressionCV.html) - [3.2.4.1.6. `sklearn.linear_model`.MultiTaskElasticNetCV](modules/generated/sklearn.linear_model.MultiTaskElasticNetCV.html) - [3.2.4.1.7. `sklearn.linear_model`.MultiTaskLassoCV](modules/generated/sklearn.linear_model.MultiTaskLassoCV.html) - [3.2.4.1.8. `sklearn.linear_model`.OrthogonalMatchingPursuitCV](modules/generated/sklearn.linear_model.OrthogonalMatchingPursuitCV.html) - [3.2.4.1.8.1. Examples using `sklearn.linear_model.OrthogonalMatchingPursuitCV`](modules/generated/sklearn.linear_model.OrthogonalMatchingPursuitCV.html#examples-using-sklearn-linear-model-orthogonalmatchingpursuitcv) - [3.2.4.1.9. `sklearn.linear_model`.RidgeCV](modules/generated/sklearn.linear_model.RidgeCV.html) - [3.2.4.1.9.1. Examples using `sklearn.linear_model.RidgeCV`](modules/generated/sklearn.linear_model.RidgeCV.html#examples-using-sklearn-linear-model-ridgecv) - [3.2.4.1.10. `sklearn.linear_model`.RidgeClassifierCV](modules/generated/sklearn.linear_model.RidgeClassifierCV.html) - [3.2.4.2. 信息標準](modules/grid_search.html#id23) - [3.2.4.2.1. `sklearn.linear_model`.LassoLarsIC](modules/generated/sklearn.linear_model.LassoLarsIC.html) - [3.2.4.2.1.1. Examples using `sklearn.linear_model.LassoLarsIC`](modules/generated/sklearn.linear_model.LassoLarsIC.html#examples-using-sklearn-linear-model-lassolarsic) - [3.2.4.3. 出袋估計](modules/grid_search.html#out-of-bag) - [3.2.4.3.1. `sklearn.ensemble`.RandomForestClassifier](modules/generated/sklearn.ensemble.RandomForestClassifier.html) - [3.2.4.3.1.1. Examples using `sklearn.ensemble.RandomForestClassifier`](modules/generated/sklearn.ensemble.RandomForestClassifier.html#examples-using-sklearn-ensemble-randomforestclassifier) - [3.2.4.3.2. `sklearn.ensemble`.RandomForestRegressor](modules/generated/sklearn.ensemble.RandomForestRegressor.html) - [3.2.4.3.2.1. Examples using `sklearn.ensemble.RandomForestRegressor`](modules/generated/sklearn.ensemble.RandomForestRegressor.html#examples-using-sklearn-ensemble-randomforestregressor) - [3.2.4.3.3. `sklearn.ensemble`.ExtraTreesClassifier](modules/generated/sklearn.ensemble.ExtraTreesClassifier.html) - [3.2.4.3.3.1. Examples using `sklearn.ensemble.ExtraTreesClassifier`](modules/generated/sklearn.ensemble.ExtraTreesClassifier.html#examples-using-sklearn-ensemble-extratreesclassifier) - [3.2.4.3.4. `sklearn.ensemble`.ExtraTreesRegressor](modules/generated/sklearn.ensemble.ExtraTreesRegressor.html) - [3.2.4.3.4.1. Examples using `sklearn.ensemble.ExtraTreesRegressor`](modules/generated/sklearn.ensemble.ExtraTreesRegressor.html#examples-using-sklearn-ensemble-extratreesregressor) - [3.2.4.3.5. `sklearn.ensemble`.GradientBoostingClassifier](modules/generated/sklearn.ensemble.GradientBoostingClassifier.html) - [3.2.4.3.5.1. Examples using `sklearn.ensemble.GradientBoostingClassifier`](modules/generated/sklearn.ensemble.GradientBoostingClassifier.html#examples-using-sklearn-ensemble-gradientboostingclassifier) - [3.2.4.3.6. `sklearn.ensemble`.GradientBoostingRegressor](modules/generated/sklearn.ensemble.GradientBoostingRegressor.html) - [3.2.4.3.6.1. Examples using `sklearn.ensemble.GradientBoostingRegressor`](modules/generated/sklearn.ensemble.GradientBoostingRegressor.html#examples-using-sklearn-ensemble-gradientboostingregressor) - [3.3. 模型評估: 量化預測的質量](modules/model_evaluation.html) - [3.3.1. `scoring` 參數: 定義模型評估規則](modules/model_evaluation.html#scoring) - [3.3.1.1. 常見場景: 預定義值](modules/model_evaluation.html#id2) - [3.3.1.2. 根據 metric 函數定義您的評分策略](modules/model_evaluation.html#metric) - [3.3.1.3. 實現自己的記分對象](modules/model_evaluation.html#diy-scoring) - [3.3.1.4. 使用多個指數評估](modules/model_evaluation.html#multimetric-scoring) - [3.3.2. 分類指標](modules/model_evaluation.html#classification-metrics) - [3.3.2.1. 從二分到多分類和 multilabel](modules/model_evaluation.html#multilabel) - [3.3.2.2. 精確度得分](modules/model_evaluation.html#accuracy-score) - [3.3.2.3. Cohen’s kappa](modules/model_evaluation.html#cohen-s-kappa) - [3.3.2.4. 混淆矩陣](modules/model_evaluation.html#confusion-matrix) - [3.3.2.5. 分類報告](modules/model_evaluation.html#classification-report) - [3.3.2.6. 漢明損失](modules/model_evaluation.html#hamming-loss) - [3.3.2.7. Jaccard 相似系數 score](modules/model_evaluation.html#jaccard-score) - [3.3.2.8. 精準，召回和 F-measures](modules/model_evaluation.html#f-measures) - [3.3.2.8.1. 二分類](modules/model_evaluation.html#id14) - [3.3.2.8.2. 多類和多標簽分類](modules/model_evaluation.html#id15) - [3.3.2.9. Hinge loss](modules/model_evaluation.html#hinge-loss) - [3.3.2.10. Log 損失](modules/model_evaluation.html#log) - [3.3.2.11. 馬修斯相關系數](modules/model_evaluation.html#matthews-corrcoef) - [3.3.2.12. Receiver operating characteristic (ROC)](modules/model_evaluation.html#receiver-operating-characteristic-roc) - [3.3.2.13. 零一損失](modules/model_evaluation.html#zero-one-loss) - [3.3.2.14. Brier 分數損失](modules/model_evaluation.html#brier) - [3.3.3. 多標簽排名指標](modules/model_evaluation.html#multilabel-ranking-metrics) - [3.3.3.1. 覆蓋誤差](modules/model_evaluation.html#coverage-error) - [3.3.3.2. 標簽排名平均精度](modules/model_evaluation.html#label-ranking-average-precision) - [3.3.3.3. 排序損失](modules/model_evaluation.html#label-ranking-loss) - [3.3.4. 回歸指標](modules/model_evaluation.html#regression-metrics) - [3.3.4.1. 解釋方差得分](modules/model_evaluation.html#explained-variance-score) - [3.3.4.2. 平均絕對誤差](modules/model_evaluation.html#mean-absolute-error) - [3.3.4.3. 均方誤差](modules/model_evaluation.html#mean-squared-error) - [3.3.4.4. 均方誤差對數](modules/model_evaluation.html#mean-squared-log-error) - [3.3.4.5. 中位絕對誤差](modules/model_evaluation.html#median-absolute-error) - [3.3.4.6. R2 score, 可決系數](modules/model_evaluation.html#r2-score) - [3.3.5. 聚類指標](modules/model_evaluation.html#clustering-metrics) - [3.3.6. 虛擬估計](modules/model_evaluation.html#dummy-estimators) - [3.4. 模型持久化](modules/model_persistence.html) - [3.4.1. 持久化示例](modules/model_persistence.html#id2) - [3.4.2. 安全性和可維護性的局限性](modules/model_persistence.html#persistence-limitations) - [3.5. 驗證曲線: 繪制分數以評估模型](modules/learning_curve.html) - [3.5.1. 驗證曲線](modules/learning_curve.html#validation-curve) - [3.5.2. 學習曲線](modules/learning_curve.html#learning-curve)