首页> 中文学位 >基于变参神经动力学方法的分形生成方法研究
【6h】

基于变参神经动力学方法的分形生成方法研究

代理获取

目录

声明

Chapter 1 Introduction

1.1 Fractals

1.1.1 Fractals in Nature

1.1.2 Fractals Geometry

1.1.3 Fractals Properties

1.1.4 Fractals Applications

1.1.5 Conclusions

1.2 Research Objectives

1.3 Methodology

1.4 Dissertation Structure

Chapter 2 Fixed-Parameter Neural Dynamics Fractals

2.1 Model of Fixed Parameter Neural Dynamics

2.1.1 Relate to Newton Iterations

2.2 Fractals Generating

2.2.1 Newton Fractals

2.2.2 FP-DTCVND Fractals

2.3 Varying-Parameters Convergent-Differential Neural-Network (VP-CDNN)

2.3.1 VP-CDNN Model

2.3.2 Comparison with FT-ZNN

Chapter 3 Exponential-Type Varying-Parameter Neural Dynamics Fractals

3.1 Model of the Exponential-Type Varying-Parameter Neural Dynamics

3.1.1 Varying-Parameters Discrete Time Complex-Valued Neural (YP-DTCVND)

3.1.2 Model for Solving Static Nonlinear Equations

3.2 Exponential-Type VP-DTCVND Fractals Generating

3.2.2 Fractals via f(z,t)=sin(z)+exp(z)-sin(t)-i·t=0

3.2.3 Fractals via f(z,t)=exp(z)7+i·cos(t)=0

3.3 Comparing between Fractals of FP-DTCVND and the Exponential-Type VP-DTCVND

3.3.1 Comparison Considering the Linear Activation Function

3.3.2 Comparison Considering the Power-sum Activation Function

3.3.3 Comparison Considering the Power-sigmoid Activation Function

3.3.4 Comparison Considering the Hyperbolic-sine Activation Function

3.3.5 Comparison between the Activation Function Performance

Chapter 4 Power-Type Varying-Parameter Neural Dynamics Fractals

4.1 Power-Type VP-DTCVND Model

4.1.1 Different Activation Functions

4.1.2 Model for Solving Static Nonlinear Equations

4.1.3 Comparison between Convolution Models

4.2 Power-Type VP-DTCVND Fractals Generating

4.2.2 Fractals via f(z)=cos(2s)=0

4.2.3 Fractals via f(z)=z3exp(z3+i)=0

4.2.4 Fractals via f(z,t)=z(t)3+exp(z)3+i·t=0

4.2.5 Fractals via f(z,t)=exp(z)+z(t)2-exp(t)-i·t=0

4.2.6 Fractals via f(z,t)=sin(z)+z(t)2-sin(t)-i·t=0

4.3 Comparing between Fractals of FP-DTCVND and the power-Type VP-DTCVND

4.3.1 Comparison Considering the Linear Activation Function

4.3.2 Comparison Considering the Power-sum Activation Function

4.3.3 Comparison Considering the Power-sigmoid Activation Function

4.3.4 Comparison Considering the Hyperbolic-sine Activation Function

4.3.5 Comparison between the Activation Function Performance

Chapter 5 Conclusion

5.1 Contributions Highlight

5.2 Future Work

参考文献

攻读硕士学位期间取得的研究成果

致谢

答辩决议书

展开▼

著录项

  • 作者

    阿克林Akram Ahmad;

  • 作者单位

    华南理工大学;

  • 授予单位 华南理工大学;
  • 学科 电气与计算机工程
  • 授予学位 硕士
  • 导师姓名 张智军;
  • 年度 2019
  • 页码
  • 总页数
  • 原文格式 PDF
  • 正文语种 chi
  • 中图分类 TN9;
  • 关键词

相似文献

  • 中文文献
  • 外文文献
  • 专利
代理获取

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号