12 papers to understand QA system with Deep Learning

由于最近入手NLP任务,需要看一些paper,本文对最近两周看的paper做个总结,适用于有deep learning背景,希望了解NLP应用的同学,主要针对NLP方向: 问答系统(QA)和翻译(Machine Translation)。本文提到的12篇paper比较有代表性,这里感谢总理和江哥提供部分参考paper和指导帮助。

论文列表:(其中QA为Question Answer的缩写)

  1. Neural Machine Translation by Jointly Learning to Align and Translate
    任务: 机器翻译
    关键词:attention BiRNN
    中心思想: English -> encoder -> decoder -> Chinese。其中encoder一般是一个RNN,读入一个词序列,输出一个表示该句话的vector;decoder一般也是一个RNN,输入该句话的表示vector,再以序列输出,每个时刻预测下一个词 yt

    常用优化目标:令 p(y) 最大,其中






    c是encoder输出的原句vector表示,
    st 是decoder RNN的 hidden state,
    yt1 t1 时刻预测的翻译词,
    g 是非线性函数。

    i.e., 基于{上一时刻预测词,当前decoder状态,输入句子(待翻译句子)的encoder vector表示} 确定当前时刻输出词。

    方法:本文中,
    encoder: 一个双向RNN,从前到后,从后往前各读一遍输入序列
    decoder: encoder的c变成了 ci

  • 22
    点赞
  • 47
    收藏
    觉得还不错? 一键收藏
  • 22
    评论
As an AI language model, I can suggest the following steps to learn deep learning: 1. Learn the basics of machine learning: Before diving into deep learning, it is essential to understand the basics of machine learning, including supervised and unsupervised learning, regression, clustering, and classification. 2. Learn programming languages: To work with deep learning, one needs to have a strong foundation in programming languages such as Python and R, as most of the deep learning libraries are written in these languages. 3. Understand the mathematics behind deep learning: Deep learning involves a lot of math, including linear algebra, calculus, and probability. Understanding these concepts will help you better understand the algorithms used in deep learning. 4. Choose a deep learning framework: Popular deep learning frameworks include Tensorflow, Keras, PyTorch, and Caffe. Choose one and learn it. 5. Practice with datasets: Work with datasets to understand how deep learning works in practice. Kaggle is a great platform to get started with real-world datasets. 6. Read research papers: Read research papers to stay up-to-date with the latest advancements in deep learning. 7. Join communities: Join online communities such as Reddit, Discord, or GitHub to connect with other deep learning enthusiasts and learn from them. 8. Build projects: Building projects is the best way to learn deep learning. Start with simple projects and gradually move on to more complex ones. Remember, deep learning is a vast field, and it takes time and effort to master it. Keep practicing, and you will get there.

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论 22
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值