知识卡片
张量阶数是轴的数量
张量以多个索引轴组织数值:标量无轴,向量一轴,矩阵两轴;用户—地点—时间是三阶。阶数说明有多少类索引,不是每轴的长度,也不直接说明模型复杂度或秩。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
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
Scalars are interpreted as 0-order tensors, and vectors are represented as first-order tensors. Second-order tensors represent matrices. Now POI recommendation came up with third-order or higher dimensional tensors. In the case of a third-order tensor, the fibers [18] are divided in three ways, such as a row, column, and tube sliced in frontal, lateral, and horizontal patterns. The elements of tensors are usually of the
参考来源
- 位置:《Artificial Intelligence and Data Science in Recommendation System: Current Trends, Technologies, and Applications》第11章《张量分解的兴趣点推荐》(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-C11.html)
- 结论依据:摘录中关于张量阶数是轴的数量的定义、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:Scalars are interpreted as 0-order tensors, and vectors are represented as first-order tensors. Second-order tensors represent matrices. Now POI recommendation came up with third-order or higher dimensional tensors. In the case of a third-order tensor, the fibers [18] are divided in three ways, such as a row, column, and tube sliced in frontal, lateral, and horizontal patterns. The elements of tensors are usually of the