摘要
随着信息化技术的迅速发展,社会各个行业对信息化技术的依赖与日俱深,其中一个重要的场景就是将用户的手写信息识别到计算机系统中,而在这一场景中一个常见识别内容就是识别手写数字。传统方法难以准确识别手写数字,深度学习则计算量较大,因此提出使用机器学习中决策树模型来分析手写数字。实验结果表明,所提方法能在保持计算量较小的前提下精确、快速地识别出各种手写数字,可作为手写数字的有效识别手段。
With the rapid development of information technology in recent years, all sectors of society rely on information technology more than ever before. One of the important scenes is to recognize the user＇s handwritten information into the computer system. In this scene, a common recognition content is to recognize handwritten digits, and traditional methods are difficult to recognize handwritten digits accurately, while the depth learning requires a lot of computation. Therefore, a decision tree in machine learning is proposed to analyze handwritten digits. Experimental results show that the proposed method can recognize various handwritten digits accurately and quickly with less computation. It can be used as an effective recognition method for handwritten digits.
作者
赵力衡
ZHAO Li-heng(Jincheng College, Sichuan University, Chengdu Sichuan 611731, Chin)
出处
《软件》
2018年第3期90-94,共5页
Software
基金
四川省教育厅重大培育项目(18CZ0047)
关键词
大数据
机器学习
决策树
手写数字
Big data
Machine learning
Decision tree
Handwriting digital
