当前位置:网站首页>Novice entry depth learning | 3-6: optimizer optimizers
Novice entry depth learning | 3-6: optimizer optimizers
2022-07-06 00:41:00 【Classmate K】
- Running environment :python3
- author :K Students
- Select columns :《 Deep learning 100 example 》
- From column :《 Novice introduction and deep learning 》
- Recommendation column :《Matplotlib course 》
- 🧿 Excellent column :《Python introduction 100 topic 》
The optimizer is in the process of deep learning back propagation , Guide the loss function ( Objective function ) The parameters of are updated in the correct direction and the appropriate size , Make the updated parameters let the loss function ( Objective function ) The value keeps approaching the global minimum .
I wrote a lot of things , But I feel that talking too much will be counterproductive ,
The choice of model optimization algorithm is directly related to the performance of the final model . Sometimes it doesn't work , It's not necessarily about features or model design , It's probably the problem of optimizing the algorithm .
Deep learning optimization algorithm has probably experienced SGD -> SGDM -> NAG ->Adagrad -> A
边栏推荐
- MIT博士论文 | 使用神经符号学习的鲁棒可靠智能系统
- Synchronized and reentrantlock
- Starting from 1.5, build a micro Service Framework - call chain tracking traceid
- Reading notes of the beauty of programming
- LeetCode 6005. The minimum operand to make an array an alternating array
- notepad++正则表达式替换字符串
- Solve the problem of reading Chinese garbled code in sqlserver connection database
- NLP generation model 2017: Why are those in transformer
- 如何利用Flutter框架开发运行小程序
- Spark AQE
猜你喜欢
Hudi of data Lake (2): Hudi compilation
数据分析思维分析方法和业务知识——分析方法(二)
多线程与高并发(8)—— 从CountDownLatch总结AQS共享锁(三周年打卡)
MDK debug时设置数据实时更新
LeetCode 1189. Maximum number of "balloons"
Basic introduction and source code analysis of webrtc threads
小程序技术优势与产业互联网相结合的分析
KDD 2022 | 脑电AI助力癫痫疾病诊断
The relationship between FPGA internal hardware structure and code
时间戳的拓展及应用实例
随机推荐
LeetCode 1189. Maximum number of "balloons"
STM32 key chattering elimination - entry state machine thinking
Priority queue (heap)
Go learning --- structure to map[string]interface{}
Power query data format conversion, Split Merge extraction, delete duplicates, delete errors, transpose and reverse, perspective and reverse perspective
Classic CTF topic about FTP protocol
uniapp开发,打包成H5部署到服务器
FPGA内部硬件结构与代码的关系
SQLServer连接数据库读取中文乱码问题解决
[groovy] compile time metaprogramming (compile time method injection | method injection using buildfromspec, buildfromstring, buildfromcode)
Comment faire votre propre robot
[Chongqing Guangdong education] Chongqing Engineering Vocational and Technical College
STM32按键消抖——入门状态机思维
Natural language processing (NLP) - third party Library (Toolkit):allenlp [library for building various NLP models; based on pytorch]
Leetcode 44 Wildcard matching (2022.02.13)
电机的简介
Browser local storage
[groovy] compile time meta programming (compile time method interception | method interception in myasttransformation visit method)
Extracting profile data from profile measurement
State mode design procedure: Heroes in the game can rest, defend, attack normally and attack skills according to different physical strength values.