当前位置:网站首页>荒野觅踪---寻找迭代次数
荒野觅踪---寻找迭代次数
2022-07-27 10:26:00 【黑榆】
按照迭代次数和熵之间反比关系的假设

(A,B)---m*n*k---(1,0)(0,1)
可以合理的推测两个分类对象之间的熵取决于点的数值大小和分布,这次就从数值大小的角度验证迭代次数和熵之间的关系。
(0,3)---m*n*k---(1,0)(0,1)
用神经网络分类mnist的0和3,用间隔取点的办法把图片缩小为9*9,让0就是原始数据集的0。mnist 3随机的保留70%的点,并在归一化的操作中增加一个系数d。让d分别等于1,0.9,0.8,…,0.1.比如如果d=0.1就意味这3中的所有点的最大值就是0.1,相当于归0.1化。而0都按照正常的方式归一化。因此点的分布完全相同,而迭代次数的差异仅仅取决于数值大小。
考虑两个粒子n和m,n的运动范围是0-1,另一个粒子m运动范围是0-0.1,因为n粒子的运动区间包含了m粒子的运动区间,所以m粒子可能的状态数应该小于n粒子可能的运动状态数量。因此n粒子的熵应该大于m粒子的熵。因此系数d越大熵越大。
现在不管形态0和3到底是如何相互作用的,把0和3考虑成一个复合的整体,这个整体在点的位置关系没有变化的情况下,点的数值大小变了。系数d值对0和3的复合体的影响应该和对n和m的影响是一致的。
因此有理由假设随着系数d的减小0和3的复合体的熵减小,迭代次数增加。
让收敛误差为1e-4到1e-5,每个收敛误差统计199次,取平均值。得到表格
*d:1 | 0.9 | 0.8 | 0.7 | 0.6 | 0.5 | 0.4 | 0.3 | 0.2 | 0.1 | |
δ | 迭代次数n | 迭代次数n | 迭代次数n | 迭代次数n | 迭代次数n | 迭代次数n | 迭代次数n | 迭代次数n | 迭代次数n | 迭代次数n |
1.00E-04 | 2376.150754 | 2800.271357 | 2626.100503 | 3907.909548 | 3742 | 5205.105528 | 6087.638191 | 10873.94975 | 20774.23116 | 29083.03015 |
9.00E-05 | 2524.733668 | 2889.929648 | 2886.552764 | 4186.140704 | 3742 | 5209.59799 | 6454.221106 | 11297.8995 | 22375.69849 | 32312.17588 |
8.00E-05 | 2667.58794 | 2903.487437 | 2983.366834 | 4295.537688 | 3842.110553 | 5210 | 7736.271357 | 12169.38693 | 23928.75377 | 36107.54271 |
7.00E-05 | 2739.507538 | 2920.79397 | 3003.557789 | 4363.648241 | 4779.698492 | 5306.78392 | 8575.718593 | 14153.8593 | 24540.13065 | 41005.40201 |
6.00E-05 | 2953.798995 | 2961.738693 | 3026.753769 | 4597.005025 | 5197.949749 | 5557.688442 | 8658.653266 | 14807.96985 | 27646.0804 | 47307.65327 |
5.00E-05 | 3101.477387 | 2985.085427 | 3089.326633 | 4791.095477 | 5375.537688 | 5684.693467 | 8678 | 15468.83417 | 31340.38191 | 57505.47236 |
4.00E-05 | 3389.98995 | 3038.01005 | 3735.59799 | 4934.643216 | 5957.306533 | 7151.969849 | 8743.668342 | 16236.35176 | 34356.0402 | 69970.20603 |
3.00E-05 | 3531.025126 | 3569.427136 | 4368.201005 | 5044 | 6266 | 7600.562814 | 10017.12563 | 18448.62312 | 42366.81407 | 93417.52764 |
2.00E-05 | 3848.40201 | 4187.688442 | 4744 | 5140.020101 | 6266 | 9674.452261 | 12889.20603 | 23425.88945 | 52728.90452 | 134846.8593 |
1.00E-05 | 4212.81407 | 4744 | 5361.778894 | 5970 | 7748.281407 | 10620 | 14199.10553 | 27930.41206 | 74308.67337 | 234247.6332 |
比如2739.507538,这个值表明在d=1,收敛误差为7e-5的情况下收敛199次的平均值为2739.507538. 选择收敛误差分别等于3e-5,2e-5,1e-5的几组画成图

可以看到随着d的减小迭代次数增加,所以这个实验现象和假设符合的很好。
这个实验验证了迭代次数和熵之间成反比的假设,也某种程度上表明了把完成收敛的0和3看成是一个复合体的可能,表明了对于两个二维形态的熵可以体现一种非递进的二维形态表象,同时也有一个一维的连续的线性递进表象。
或者点的位置关系决定了形态复合体的一种二维非递进表象,而数值的大小决定了形态复合体的一维连续表象。
边栏推荐
- Analysis of new communication security risks brought by quantum computer and Countermeasures
- Your appearance is amazing! Two JSON visualization tools are recommended for use with swagger. It's really fragrant
- It is thought-provoking: is syntax really important? Qiu Xipeng group proposed a powerful baseline for aspect based emotional analysis
- 服务器访问速度
- It is thought-provoking: is syntax really important? Qiu Xipeng group proposed a powerful baseline for aspect based emotional analysis
- Overview of radar communication integrated waveform design
- Advanced operation of MySQL data table
- The permission problem of Oracle operating openldap
- 厉害了!VMware ESXi安装记录,附下载
- Edata base, a secondary development project based on spark packaging, is introduced
猜你喜欢

异构计算技术分析

搭建 Samba 服务

ASP.NET Core依赖注入之旅:1.理论概念

北京公示儿童鞋抽查 8组批产品不合格琪尔特登榜

颜值爆表!推荐两款JSON可视化工具,配合Swagger使用真香

基于Spark封装的二次开发工程edata-base,介绍

招聘顶尖人才!旷视科技“MegEagle创视者计划”正式启动

MySQL master-slave architecture, read-write separation, and high availability architecture

The core concept and fast practice of shardingsphere

Metaspolit
随机推荐
How to smooth the online and offline of Web Services
Application scenarios, key technologies and network architecture of communication perception integration
异构计算技术分析
华硕无双,这可能是屏幕最好的平价高刷轻薄笔记本
Integrated design of communication perception based on CSI: problems, challenges and Prospects
kgdb调试内核无法执行断点及kdb-22:Permisson denied
Pyqt5 rapid development and practice 4.2 QWidget
Family Trivia
File upload vulnerability related
Using skills of word
antd table+checkbox 默认值显示
antd中table hover上去的背景色样式修改
No identifier specified for entity solution
Awesome! VMware esxi installation record, with download
Substr and substring function usage in SQL
Alibaba mailbox web login turn processing
Apache22 and opencms9.5 integrated configuration
Li Hongyi_ Machine learning_ Assignment 4 (detailed explanation)_ HW4 Classify the speakers
Open source project - taier1.2 release, new workflow, tenant binding simplification and other functions
Error in nodejs: getaddrinfo enotfound localhost