Scadao 发表于 2019-1-17 14:16:58

解决几乎任何机器学习问题(完整翻译)

解决几乎任何机器学习问题(完整翻译)
https://img-blog.csdn.net/20180116090242952?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvZ2FtbWFn/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast

Scadao 发表于 2019-2-13 13:58:06

本帖最后由 Scadao 于 2019-9-18 21:32 编辑

https://img-blog.csdn.net/2018101515170874?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2xhb2JhaTEwMTU=/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70第一步是确定问题
https://img-blog.csdn.net/20180116090332653?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvZ2FtbWFn/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast

下一步是识别数据中的不同变量
下一步（或中级）的主要步骤是模型选择+超参数优化
https://img-blog.csdn.net/20180116091349831?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvZ2FtbWFn/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast

再次强调记住保存转换状态数据
https://img-blog.csdn.net/20180116091442151?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvZ2FtbWFn/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast

并将它们分别应用于验证集
https://img-blog.csdn.net/20180116091503294?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvZ2FtbWFn/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast



页: [1]
查看完整版本: 解决几乎任何机器学习问题(完整翻译)