3つの要点 ✔️ 継続学習と言語処理の融合 ✔️ 構文と意味を分離することで破滅的忘却を防止 ✔️ 既存手法と比較し圧倒的な性能 Compositional Language Continual Learning written by Yuanpeng Li Liang Zhao Kenneth Church Mohamed Elhoseiny (Submitted on 26 Sep 2019 (modified: 11 Mar 2020)) Comments: accepted by ICLR 2020 Conference Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML) はじめに 機械学習における課題の一つとして、破滅的忘却(Catastrophic Forgetting)と呼ばれる問題があります。これは、
![継続学習+言語処理の最先端!Compositionalityを用いたSequence-to-Sequenceモデル!](https://cdn-ak-scissors.b.st-hatena.com/image/square/c7108edc1816a9c10d7bed130aacd91ee1c43626/height=288;version=1;width=512/https%3A%2F%2Faisholar.s3.ap-northeast-1.amazonaws.com%2Fposts%2FApril2020%2F%25E8%2587%25AA%25E7%2584%25B6%25E8%25A8%2580%25E8%25AA%259E%25E3%2581%25AF%25E3%2581%2593%25E3%2581%2593%25E3%2581%25BE%25E3%2581%25A6%25E3%2582%2599%25E3%2581%258D%25E3%2581%25A6%25E3%2581%2584%25E3%2582%258B-1024x576.png)