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：
构造性机器学习方法——覆盖算法学习速度快、复杂度低、可解释性强，能有效地解决有导师学习问题，并取得了很好的效果，但构造神经元的权值即取新覆盖中心时通常人为地给定一个准则，并未遵循样本的分布特征求得最优解．由此采用佳点集理论求取覆盖中心，以改进覆盖算法．针对大规模或动态数据集的分类问题。将构造性覆盖方法与增量学习的思想相结合，提出了构造性覆盖方法的增量学习算法。该算法利用改进的覆盖算法作为基础学习器，通过连续地对新增样本进行测试而反复不断地提炼已有模型，体现了对样本的“渐近式”学习,对标准数据集的实验结果表明，这种增量学习算法是有效的。
英文摘要：
The structural machine learning method covering algorithm possesses faster speed, lower complexity, stronger interpretability and higher precision. It solved the problem of the supervised learning effectively and achieved the favorable performance. But the construction of the weight of the neurons-for new center of sphere domain is usually given to a man-made criteria, did not follow the distribution of samples to achieve the optimal solution. So we improve covering the algorithm inspired by Good-Point-Set. The structural covering method and increment learning are combined, and an incremental learning algorithm for structural covering method is proposed to classify large-scale or dynamic data set. This incremental learning algorithm for structural covering method uses improved covering algorithm as base classifiers and refines the existing model through testing some new samples repeated, which reflects the ＂progressive＂ study of samples. The experimental results with the standard dataset show the validity of the novel incremental learning algorithm. Moreover, the incremental learning algorithm provides an idea or a way for implementing classify in dynamic learning. It can also help us to solve many practical problems.
同期刊论文项目
同项目期刊论文
期刊信息
位置：