Problem: time series forecasting

Challenge: forecasting for non-stationary signals and multiple future steps prediction

?? how to deal with non-stationary datasets??

Introduction

one-step prediction problem VS multi-step prediction;

multi-step forecasting requires to accurately describe time series evolution.

limitation of the euclidean loss(MSE): in non-stationary context;

PP: Shape and time distortion loss for training deep time series forecasting models的更多相关文章

  1. Training (deep) Neural Networks Part: 1

    Training (deep) Neural Networks Part: 1 Nowadays training deep learning models have become extremely ...

  2. Training Deep Neural Networks

    http://handong1587.github.io/deep_learning/2015/10/09/training-dnn.html  //转载于 Training Deep Neural ...

  3. PP: Multi-Horizon Time Series Forecasting with Temporal Attention Learning

    Problem: multi-horizon probabilistic forecasting tasks; Propose an end-to-end framework for multi-ho ...

  4. a Javascript library for training Deep Learning models

    w强化算法和数学,来迎接机器学习.神经网络. http://cs.stanford.edu/people/karpathy/convnetjs/ ConvNetJS is a Javascript l ...

  5. PP: Think globally, act locally: A deep neural network approach to high-dimensional time series forecasting

    Problem: high-dimensional time series forecasting ?? what is "high-dimensional" time serie ...

  6. [Xavier] Understanding the difficulty of training deep feedforward neural networks

    目录 概 主要内容 Glorot X, Bengio Y. Understanding the difficulty of training deep feedforward neural netwo ...

  7. 论文翻译：BinaryConnect: Training Deep Neural Networks with binary weights during propagations

    目录 摘要 1.引言 2.BinaryConnect 2.1 +1 or -1 2.2确定性与随机性二值化 2.3 Propagations vs updates 2.4 Clipping 2.5 A ...

  8. 论文翻译：BinaryNet: Training Deep Neural Networks with Weights and Activations Constrained to +1 or −1

    目录 摘要 引言 1.BinaryNet 符号函数 梯度计算和累积 通过离散化传播梯度 一些有用的成分 算法1 使用BinaryNet训练DNN 算法2 批量标准化转换(Ioffe和Szegedy,2 ...

  9. Xavier——Understanding the difficulty of training deep feedforward neural networks

    1. 摘要 本文尝试解释为什么在深度的神经网络中随机初始化会让梯度下降表现很差,并且在此基础上来帮助设计更好的算法. 作者发现 sigmoid 函数不适合深度网络,在这种情况下,随机初始化参数会让较深 ...

随机推荐

  1. ES6 - 基础学习(4): 模板字符串和字符串新增方法

    模板字符串 模板字符串:我理解为将字符串格式化.模板化,将字符串加强处理,此处的模板有动词的意思. 字符串模板基本格式: `xxxxxx`(前后都用反引号[tab键上面按键]引起来).除了作为普通字符 ...

  2. Docker 网络原理

    引言 学习docker网络,可以带着下面两个问题来探讨 容器之间可以相互访问的原理 容器暴露端口后,通过宿主机访问到容器内应用,并且对于访问端而言不用感知容器存在的原理 Docker 本身的技术依赖L ...

  3. Q&A in 2018 - Q2

    How to zip a file for Windows? 压缩一个文件: makecab c:/file_name.txt c:/file_name.zip 解压一个文件: expand c:/f ...

  4. AGC011-E Increasing Numbers

    题意 给定一个数\(n\),\(n≤10^{500,000}\),问\(n\)最少可以拆分成几个不降数的和.一个不降数是在十进制位下,从高位往低位看,每个数都不会比高位的数更小的数 做法 不降数可以拆 ...

  5. Docker--数据管理之Volumes

    前言:我们知道docker容器内产生或修改的数据仅在该容器内有效,即容器关闭,其之前产生或修改的数据也就删除了,这明显不能满足我们大多数场景的需求.当然这只是默认,docker为我们提供了多种保存数据 ...

  6. vue(七)--监听属性（watch）

    1.watch:用来监听每一个属性的变化 2.watch这个对象里面都是函数,函数的名称是data中的属性名称,watch中的函数不需要调用 3.当属性发生改变那么就会触发watch函数,每个函数都会 ...

  7. Centos 7 firewall的防火墙的规则

    这是官方文档: http://www.firewalld.org/documentation/man-pages/firewall-cmd.html 想使用iptables的规则,firewall也可 ...

  8. A - Kvass and the Fair Nut 二分

    The Fair Nut likes kvass very much. On his birthday parents presented him nn kegs of kvass. There ar ...

  9. [译]C# 7系列，Part 10: Span<T> and universal memory management Span<T>和统一内存管理

    原文:https://blogs.msdn.microsoft.com/mazhou/2018/03/25/c-7-series-part-10-spant-and-universal-memory- ...

  10. markdown转成word或者pdf

    利用typora软件 1.登陆官网下载软件 官网地址:https://typora.io/ 点击download 根据自己的电脑下载64位或者32位 2.安装软件 安装界面如下: 3.转换 3.1首先 ...