|
COVER ABSTRACT(CHINESE) ABSTRACT(ENGLISH) ACKNOWLEDGMENT CONTENTS TABLE CAPTIONS FIGURE CAPTIONS CHAPTER 1 INTRODUCTION 1.1 BACKGROUND 1.2 ARTIFICIAL NEURAL NETWORK IMPLEMENTATIONS 1.3 ANALOG IMPLEMENTATIONS OF NEURAL SYSTEMS PROBLEMS AND SOLUTIONS 1.4 APPLICATIONS 1.5 ORGANIZATION OF THIS THESIS CHAPTER 2 CMOS CURRENT-MODE NEURAL ASSOCIATIVE MEMORY DESIGN WITH ON-CHIP LEARNING 2.1 INTRODUCTION 2.2 MODEL,ARCHITECTURE,AND AMOS IMPLEMENTATION 2.3 OPRATIONAL PRINCIPLES OF THE OUTSTAR CIRCUIT 2.4 EXPERIMENTAL RESULTS 2.5 SYSTEM APPLICATIONS 2.6 SUMMARY CHAPTER 3 THE DESIGN OF ANALOG CURRENT-MODE CMOS FEEDFORWARD ON-CHIP LEARNABLE MEURAL NETWORK FOR REAL-TIME PATTERN CLASSFICATION 3.1 INTRODUCTION 3.2 RATIO MEMORY AND MODIFIED FEEDFORWARD HAMMIMG NETWORK 3.3 CMOS CIRCUIT IMPLEMENTATIONS 3.4 EXPERIMENTAL RESULTS 3.5 SUMMARY CHAPTER 4 THE MULTI-CHIP DESIGN OF ANALOG CMOS EXPANDABLE MODIDIED HAMMING NEURAL NETWORK WITH ON CHIP LEARNING AND STORAGE 4.1 INTRODUCTION 4.2 ARCHITECTURE AND CMOS CIRCUIT IMPLEMENTATION 4.3 EXPANSION METHODS AND OPERATIONAL PRINCIPLES 4.4 EXPERIMENTAL RESULTS 4.5 SUMMARY CHAPTER 5 THE DESIGN OF MODIFIED HOPFIELD NETWORK WITH LEARNING AND STORAGE CAPABILITIES 5.1 INTRODUCTION 5.2 MODEL AND ARCHITECTUR 5.3 SIMULATION RESULTS 5.4 SUMMARY CHAPTER 6 CONCLUSIONS AND FUTURE WORKS 6.1 MAIN RESULTS OF THIS THESIS 6.2 FUTURE WORKS REFERENCE VITA PUBLICATIONLIST
|