[問題] General Linear Model 做預測
很多文獻都強調 GLM 的好處是可以 numerical跟categorical variables
同時做變異數分析, 但我找不到拿 GLM 來做預測 跟回歸的比較
大部分都提到 回歸是 GLM 的special case
回歸理的categorical預測變數得轉成虛擬變數,GLM 則可以直接把level當變數的值
有人比較過兩個model在預測上有啥差別嗎 (例如優缺點或適當的統計解釋...等等)
或有文獻有兩者在預測上的比較嗎?
感謝
--
※ 發信站: 批踢踢實業坊(ptt.cc)
◆ From: 161.130.178.168
→
02/15 08:03, , 1F
02/15 08:03, 1F
→
02/15 11:22, , 2F
02/15 11:22, 2F
→
02/15 11:22, , 3F
02/15 11:22, 3F
→
02/15 11:24, , 4F
02/15 11:24, 4F
→
02/15 11:26, , 5F
02/15 11:26, 5F
→
02/15 11:27, , 6F
02/15 11:27, 6F
→
02/15 11:29, , 7F
02/15 11:29, 7F
→
02/15 11:29, , 8F
02/15 11:29, 8F
→
02/15 11:31, , 9F
02/15 11:31, 9F
→
02/15 11:31, , 10F
02/15 11:31, 10F
→
02/15 11:36, , 11F
02/15 11:36, 11F
→
02/15 11:37, , 12F
02/15 11:37, 12F
→
02/15 11:38, , 13F
02/15 11:38, 13F
→
02/15 11:38, , 14F
02/15 11:38, 14F
→
02/15 11:39, , 15F
02/15 11:39, 15F
→
02/15 11:39, , 16F
02/15 11:39, 16F
→
02/15 11:46, , 17F
02/15 11:46, 17F
→
02/15 11:47, , 18F
02/15 11:47, 18F
→
02/15 11:47, , 19F
02/15 11:47, 19F
→
02/15 11:53, , 20F
02/15 11:53, 20F
→
02/15 12:05, , 21F
02/15 12:05, 21F
→
02/15 12:06, , 22F
02/15 12:06, 22F
→
02/15 12:07, , 23F
02/15 12:07, 23F
→
02/15 12:11, , 24F
02/15 12:11, 24F
→
02/15 12:12, , 25F
02/15 12:12, 25F
→
02/15 12:13, , 26F
02/15 12:13, 26F
→
02/15 12:20, , 27F
02/15 12:20, 27F
→
02/16 23:37, , 28F
02/16 23:37, 28F