PP: Shallow RNNs: a method for accurate time-series classification on tiny devices
Problem: time series classification
shallow RNNs: the first layer splits the input sequence and runs several independent RNNs. The second layer consumes the output of the first layer to capture long dependencies.
We improve inference time over standard RNNs without compromising accuracy.
Time series -------- temporal dependencies. Sequential models such as RNN are particularly well-suited in this context.
Directly leveraging RNNs for prediction in constrained scenarios is challenging, and requires large training and inference costs.
?? how long the recurrence of RNN should be?
Each time series is divided into independent parts, and a shared RNN operates on each brick independently, thus ensuring a small model size and short recurrence.
Only has a short recurrence.
Supplementary knowledge:
1. theoretical justification
2. weak/ strong assumptions; Model flow:
- assumption/ environment ~ weak or strong
- model
- results
- evaluation ~ baseline or indicator.
3. Sequential models: RNN;
最新文章
- 算法与数据结构(十三) 冒泡排序、插入排序、希尔排序、选择排序(Swift3.0版)
- Java FtpClient 实现文件上传服务
- ModelDataExchange - Import
- Python之路【第十九章】:Django 数据库对象关系映射
- [原]__FILE__宏
- 完整成功配置wamp server小记
- android fragment 博客 学习记录
- 机器学习之单变量线性回归(Linear Regression with One Variable)
- ZeroBrane Lua脚本编辑器代码自动补全
- js过滤emoji表情符号
- Atom编辑器之加快React开发的插件汇总
- 剑指架构师系列-Redis安装与使用
- ubuntu创建新用户
- maven将依赖的jar包复制到指定位置
- FormatMessage
- 011 Linux环境下配置eclipse,以及创建maven工程
- Spring------生命周期
- tcl脚本
- gitlab、openvpn配置ldap认证
- ajax访问WebService跨域问题