知识卡片

普通读书笔记卡 · 6487.j · 生成模型:GPT-6

多源迁移需协调目标预测

理解等级:中级

本章简述TRACER以多个来源域进行迁移,并用共识正则协调目标域推荐。多源信息不是简单全部加入即可,来源之间可能存在分歧。这里只保留其共享约束的目的,未据综述简述补造损失函数或实现流程。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      TRACER’s knowledge-transferring process generates recommendations iteratively for each domain. It uses the Consensus Regularization technique to select the best recommendations for the target domain. Unlike CBT, TRACER allows two or more source domains to make recommendations without user/item overlap.

    参考来源

    - 位置:《Artificial Intelligence and Data Science in Recommendation System: Current Trends, Technologies, and Applications》第04章《多领域情境推荐的张量分解》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Artificial Intelligence and Data Science in Recommendation System: Current Trends, Technologies, and Applications/_epub-src/OEBPS/Text/9789815136746-C4.html) - 结论依据:摘录中关于多源迁移需协调目标预测的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。 - 原始内容:TRACER’s knowledge-transferring process generates recommendations iteratively for each domain. It uses the Consensus Regularization technique to select the best recommendations for the target domain. Unlike CBT, TRACER allows two or more source domains to make recommendations without user/item overlap.