欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对大规模数据集下支持向量机（SVM）训练耗时长的缺陷，以及支持向量机中核函数维数过高，采用了主成分分析法对语音特征降维，减少了核函数的阶数，降低计算复杂度，进而缩短训练模型所用时间。实验证明，该方法不仅能够缩短训练时间，而且能通过控制贡献度来保持识别率不下降。
英文摘要：
For large-scale data sets, SVM training consumes a large number of time and the dimension of the kernel functions SVM is too high. Using principal component analysis on the speech feature dimensionality, aim at reducing kernel function order and cutting down the computational complexity, thus using that method can shorten the time of training model： The experiment proves that the method is not only possible to shorten training time , but also maintains the recognition rate not decreased by controlling the contribution of the speech feature.
同期刊论文项目
同项目期刊论文
Parameter optimization and application of support vector machine based on parallel artificial fish s
Study for classification of emotional speech by using optimized frame zero crossing with peak amplit
A robust feature extraction approach based on an auditory model for classification of speech and exp
期刊信息
位置：