垃圾邮件识别
收藏海南省数据产品超市2023-12-19 更新2024-03-01 收录
下载链接:
https://www.datadex.cn/app/buyApi?id=22eed943d6614c5092e93b47c791b055
下载链接
链接失效反馈官方服务:
资源简介:
通用型能力,利用卷积神经网络训练公开邮件数据集,输入为邮件文本,输出为该邮件是垃圾邮件的概率,训练集共16000张图片,8000正例、8000负例,测试集共8000张图片。取阈值为90%,训练后的网络在测试集下准确率达到98%以上。
This general-purpose spam detection model is trained on a public email dataset using Convolutional Neural Networks (CNNs). The model takes email text as input and outputs the probability that the email is spam. The training set contains 16,000 samples, including 8,000 positive examples and 8,000 negative examples, while the test set consists of 8,000 samples. With a threshold set at 90%, the trained network achieves an accuracy of over 98% on the test set.
提供机构:
中国电信股份有限公司海南分公司-数据产品超市
创建时间:
2023-12-19
搜集汇总
数据集介绍

以上内容由遇见数据集搜集并总结生成



