作者KirimaSyaro (心懷感謝不以物喜不以己)
看板DataScience
標題[問題] Prelu是不是被時代淘汰了?
時間Wed Dec 18 21:47:09 2019
從relu 到leaky relu
再到P relu 好像很有道理 P relu本身參數可以學習
可是現在很少看到人用 大部分都直接relu 或leaky relu了
當然現在有selu這種更先進的 不過selu也沒有想像中熱門
為什麼?
--
So many people live within unhappy circumstances and yet will not take the
initiative to change their situation because they are conditioned to a life
of security, conformity, and conservation, all of which may appear to give
one peace of mind, but in reality, nothing is more damaging to the adventurous
spirit within a man than a secure future. The very basic core of a man's living
spirit is his passion for adventure. --Christopher McCandless
--
※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 140.112.16.173 (臺灣)
※ 文章網址: https://www.ptt.cc/bbs/DataScience/M.1576676836.A.BD3.html
→ numpy: 現在還有 gelu 12/18 23:37
推 yoyololicon: padolu也不錯 12/19 09:16
→ KirimaSyaro: prelu 我訓練結果有時候訓練不好 12/19 20:45
→ school4303: ru啦 12/20 00:39
推 acctouhou: 還有celu 12/20 03:09
→ acctouhou: 但簡單粗暴還是relu好用 每次用automl 都是relu最好 12/20 03:10
推 followwar: 簡單 12/27 11:24
推 sxy67230: 9012年了,還是會用sigmoid 路過 12/31 10:31
推 st1009: 要2020了,還在用softmax路過 12/31 18:22
→ ap954212: Switch好用 01/03 17:15
→ benqm300: 瑪莉歐賽車好玩 03/17 15:34