Archive: 2019/9


Few-Shot Text Classification With Distributional Signatures

本文提出了一种少样本下的元学习方法,用于文本分类,在多个数据集上取得较好效果。 paper: Work in Progress for ICLR2020


A Survey of User Simulator in Task-Oriented Dialog System

Introduction任务导向型对话系统用于帮助用户完成某个任务如查电影、找餐馆等,它一般由四个模块组成:自然语言理解模块(Natural Language Understanding, NLU)、对话状态跟踪模块(Dialog State Tracking, DST)、对话策略模块(Dialog Policy, DP)和自然语言生成模块(Natural language Generation


Meta-Learning Learning to Learn

Meta-learning, also known as “learning to learn”, intends to design models that can learn new skills or adapt to new environments rapidly with a few training examples. There are three common approaches: 1) learn an efficient distance metric (metric-based); 2) use (recurrent) network with external or internal memory (model-based); 3) optimize the model parameters explicitly for fast learning (optimization-based).