From e1db6ada821c0625008e53764369dc7809366112 Mon Sep 17 00:00:00 2001 From: hypox64 Date: Sat, 14 Dec 2019 14:42:41 +0800 Subject: [PATCH] update readme --- course_report.md | 1 + 1 file changed, 1 insertion(+) diff --git a/course_report.md b/course_report.md index a252a1f..504c7b2 100644 --- a/course_report.md +++ b/course_report.md @@ -341,6 +341,7 @@ model_svr = GridSearchCV(SVR(kernel="rbf"), | ![image](./images/0.040413_121_model_xgb.png) | ![image](./images/0.030281_138_ENet.png) | | ![image](./images/0.029568_138_bay.png) | ![image](./images/0.036021_82_GBoost.png) | | ![image](./images/0.033188_65_model_lgb.png) | ![image](./images/0.028606_114_model_svr.png) | + 通过上面的图可以发现PCA的输出维数对于传统的回归模型确实由较大的影响。不同的模型最合适的输出维数并不一样。 各种模型的最终loss如下： -- GitLab