地球资源数据云——数据资源详情
该数据集《Internet Advertisements Data Set》主要用于监督学习任务,数据形态以图像为主。 题目说明:This dataset represents a set of possible advertisements on Internet pages 任务类型:图像监督学习。 建议流程:先检查类别分布与脏样本,再用迁移学习(如 ResNet/EfficientNet)建立基线。 评估建议:使用分层切分或交叉验证,优先关注 F1、Recall、AUC 等分类指标。 可用文件:add.csv。 Context The task is to predict whether an image is an advertisement ("ad") or not ("nonad"). Content There are 1559 columns in the data.Each row in the data represent one image which is tagged as ad or nonad in the last column.column 0 to 1557 represent the actual numerical attributes of the images Acknowledgements

该数据集《Internet Advertisements Data Set》主要用于监督学习任务,数据形态以图像为主。 题目说明:This dataset represents a set of possible advertisements on Internet pages
任务类型:图像监督学习。
建议流程:先检查类别分布与脏样本,再用迁移学习(如 ResNet/EfficientNet)建立基线。
评估建议:使用分层切分或交叉验证,优先关注 F1、Recall、AUC 等分类指标。
可用文件:add.csv。
Context
The task is to predict whether an image is an advertisement ("ad") or not ("nonad").
Content
There are 1559 columns in the data.Each row in the data represent one image which is tagged as ad or nonad in the last column.column 0 to 1557 represent the actual numerical attributes of the images
Acknowledgements