摘要
笔者主要使用Orange可视化机器学习软件,结合Pandas和Glueviz等工具,以Orange数据winequality-white(白葡萄酒理化性质及质量)为处理对象,阐述了如何使用Orange软件中的工具并利用各种可比较的模型比较、分析其优劣,探讨改进的方法。实验结果证明,通过训练数据训练出的模型可以推算出测试数据中各类酒的品级。在决策树、最近邻居法、线性回归和神经网络这四种模型中,神经网络是一种较好的方法,相对于普通机器学习方法有明显的优点。
The author mainly uses Orange visual machine learning software,Pandas and Glueviz tools,taking Orange data wine-quality-white(physical and chemical properties and quality of white wine)as the processing object,expounds how to use Orange software tools,and uses various comparable models to compare and analyze their advantages and disadvantages,and discusses the improvement method.The experimental results show that the model trained by the training data can be used to calculate the grade of all kinds of wines in the test data.In the four models of decision tree,nearest neighbor method,linear regression and neural network,neural network is a better method,which has obvious advantages over ordinary machine learning methods.
作者
周晓斐
Zhou Xiaofei(Shanghai Dianji University,Shanghai 200240,China)
出处
《信息与电脑》
2018年第24期71-73,共3页
Information & Computer