-
The high-level instances H \mathcal{H} H
→ \to → 同位语
The high-level instance set H \mathcal{H} H -
The IP.
→ \to → 在空间足够的时候可以不用缩写 -
k = r c k = r_c k=rc; // k k k is the number of instance prototypes of each bag.
→ \to → 注释里面不一定完整的句子
k = r c k = r_c k=rc; // The number of instance prototypes of each bag. -
Compute the representativeness score
→ \to → 代表性已经是一个指标
Compute the representativeness -
Merge the top-k score instances into C;
→ \to → Merge … into 有点怪
C ′ \mathcal{C}' C′ = the set of top-k representative instances
C = C ∪ C ′ \mathcal{C} = \mathcal{C} \cup \mathcal{C}' C=C∪C′ -
// The HI.
→ \to → 前面已经说明它是什么
// Initialize HI. -
Merge the top-nh score prototypes
→ \to → 用什么来衡量 top?
和前面保持风格/句式一致 -
// Step 2. Execute bag embedding.
→ \to →
// Step 2. Bag embedding. -
Train a single-instance classifier F(\cdot) with the vector-label pairs (V_i, y_i);
→ \to → 应该是新空间的一个数据集
Train a classifier F(\cdot) with the new dataset {(V_i, y_i)}_{i=1}^n -
Output the final classifier F(\cdot) and the high-level instances H;
→ \to →
Output F and H;
论文修改意见 (ZengWX 20211028 伪代码小问题)
最新推荐文章于 2024-05-20 10:24:40 发布