3つの要点 ✔️ ReLU、Swishに次ぐ新たな活性化関数Mishを提案 ✔️ MNISTやCIFAR-10/100などでReLUとSwishを圧倒 ✔️ 論文筆者実装のGitHubレポは早速600以上のスターを持ち、非常に簡単に使える Mish: A Self Regularized Non-Monotonic Neural Activation Function written by Diganta Misra (Submitted on 23 Aug 2019 (v1), last revised 2 Oct 2019 (this version, v2)) Subjects: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV); Neural and Evolutionary Comp
![ついに誕生!期待の新しい活性化関数「Mish」解説](https://cdn-ak-scissors.b.st-hatena.com/image/square/c545bfd31620fe6b77b6453f1488ea92133accdd/height=288;version=1;width=512/https%3A%2F%2Faisholar.s3.ap-northeast-1.amazonaws.com%2Fmedia%2FJuly2020%2F%25E6%2596%25B0%25E3%2581%2597%25E3%2581%2584%25E6%25B4%25BB%25E6%2580%25A7%25E5%258C%2596%25E9%2596%25A2%25E6%2595%25B0%25E3%2581%25A7%25E3%2581%258D%25E3%2581%25BE%25E3%2581%2597%25E3%2581%259F-min.png)