论文部分内容阅读
支持向量机推广性能的分析是机器学习中的一项重要内容。依据可通过最小化本性支持向量个数来构造支持向量机的思路,结合稀疏学习,从贪婪方法的角度出发,提出了一种新的支持向量机,称之为贪婪支持向量机。利用UCI数据库中的乳腺癌数据集来测试贪婪支持向量机算法在平衡估计精确性和解的稀疏性方面的性能。针对设计的贪婪支持向量机,利用经验过程中的方法,得到这一类型支持向量机的推广性能。
Support vector machine to promote the performance analysis is an important part of machine learning. According to the idea that SVM can be constructed by minimizing the number of intrinsic support vectors, a new SVM is proposed from greedy approach combined with sparse learning, which is called greedy support vector machine. The breast cancer dataset in the UCI database was used to test the performance of the greedy SVM algorithm in balancing the accuracy of estimation with the sparsity of solutions. For the design of greedy support vector machines, the use of the experience of the process, get this type of support vector machine to promote performance.