Learn#: A Novel incremental learning method for text classification
作者：
Highlights：
• It faces an open environment and can continuously get new training samples.
• It can get feedback from the environment in time and adjust the learning ability.
• The training time significantly reduces without retraining the historical data.
• The predictions of multiple student models are optimal globally by RL module.
• Remove one of the most similar student models to ensure the stability of the model.
摘要
•It faces an open environment and can continuously get new training samples.•It can get feedback from the environment in time and adjust the learning ability.•The training time significantly reduces without retraining the historical data.•The predictions of multiple student models are optimal globally by RL module.•Remove one of the most similar student models to ensure the stability of the model.
论文关键词：Learn#,Incremental learning,Reinforcement learning
论文评审过程：Received 20 September 2019, Revised 20 December 2019, Accepted 8 January 2020, Available online 9 January 2020, Version of Record 21 January 2020.
论文官网地址：https://doi.org/10.1016/j.eswa.2020.113198
