摘要
支持向量机是在统计学习理论基础上发展起来的一种新的模式识别方法,在解决小样本、非线性和高维模式识别问题中表现出了许多独特优势;为实现支持向量机的实时应用,提出了基于TMS320C6711DSP芯片的支持向量机的设计和实时实现方案,给出了支持向量机在模式识别中的实时应用模型和具体的硬件电路框图,对具体的硬件接口和软件编写进行了阐述;经系统仿真,该实现方法具有较好的可靠性和快速性,可以满足支持向量机的实时实现要求。
Underlying the success of Support Vector Maehines （SVMs） , which are new methods of pattern reeognition, are mathematical foundations of statistieal learning theory. They represent speeial advantages when solving the pattern reeognition problems with finite training set, non-linearity and high dimension. For the applieation of SVMs in real time, a new scheme of support vector machines implementation in TMS320C6711DSP ehip is introdueed, including the application model of support veetor maehines in pattern recognition, the practical diagram of hardware circuit, the interface of hardware and the programming of software particularly. The result of Simulation proved that this system has a good stability and reliability, which could meet the needs of implementation of SVMs in real time.
出处
《计算机测量与控制》
CSCD
2007年第1期76-78,共3页
Computer Measurement &Control
关键词
支持向量机
统计学习理论
数字信号处理器
模式识别
support vector machines （SVMs）
statistical learning theory
digital signal processor （DSP）
pattern recognition
