知识卡片
树索引以空间划分缩小搜索范围
随机投影树按超平面递归划分空间,查询访问与自身位置有关的叶或候选分支。森林和搜索预算影响漏检与速度;几何划分不保证高维语义近邻都落同一叶,示意图也不是全部实现细节。
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
To understand how tree-based algorithms work, let’s look at Annoy, a package created by Spotify for its music recommendations. The Annoy algorithm recursively partitions the input space into smaller and smaller subspaces, using a binary tree structure. At each level of the tree, the algorithm selects a hyperplane that splits the remaining points in the subspace into two groups. Eventually, each data point is assigned
参考来源
- 位置:《Natural Language Processing in Action, Second Edition》第10章《真实世界中的大语言模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Natural Language Processing in Action, Second Edition/_epub-src/OEBPS/Text/chapter-10.html)
- 结论依据:摘录中关于树索引以空间划分缩小搜索范围的定义、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:To understand how tree-based algorithms work, let’s look at Annoy, a package created by Spotify for its music recommendations. The Annoy algorithm recursively partitions the input space into smaller and smaller subspaces, using a binary tree structure. At each level of the tree, the algorithm selects a hyperplane that splits the remaining points in the subspace into two groups. Eventually, each data point is assigned