支持向量回归的研究,作为一个新的理论和方法，支持向量机回归在训练算法和实际应用等方面有诸多值得深入探讨的课题。-Machine learning is based on the data of modern intelligence technology is an important aspect. Statistical Learning Theory (SLT) is a specialized study of the law of machine learning theory of small samples, it is built on a more solid theoretical basis for solving the problem of limited samples learning provides a unified framework also developed a new universal learning a support vector machine (sVM), a better solution to the small sample learning problems. Compared with other learning methods such as neural networks, its structure is calculated by the automatic optimization methods out, and to avoid the local minimum point, over learning and other defects.