`包裹,长文用`
`/`
`划分小节。 3. 列表用`
- `,引用用`
`,代码用`
`,重要句子用``。
4. 保留原文所有信息,不得增删改内容。
5. 不要输出``/``/``等文档级标签。
原文内容:
Detecting suspicious deviations in longitudinal medication trajectories requires distinguishing knowledge-consistent treatment variations from anomalous changes and linking local deviations to account-level decisions. We propose a knowledge-guided local-to-global (L2G) anomaly learning framework that addresses both challenges. It constructs proxy supervision at three granularities: medication occurrences, treatment episodes, and contiguous visit segments. Knowledge-consistent medication and episode replacements provide plausible negatives, while cross-category and cross-account perturbations provide suspicious proxies. A medication-multiset encoder and causal visit-sequence encoder feed a shared scoring head that produces visit-level anomaly logits. Top-\(α\) aggregation derives account-level scores from these logits, while account-level classification and class-balanced visit-localization objectives jointly optimize the same visit-level evidence. Experiments on two city cohorts compare L2G with native and proxy-adapted anomaly methods. On City-A, L2G outperforms the strongest comparators. Cross-city results show more consistent gains in localization than in account-level detection.
分析原文结构:
原文是一段连续的文本,没有明显的列表、引用或代码。可以将其划分为几个逻辑段落,或者直接用一个``包裹,或者用`
`/`
`划分小节。
为了符合“长文用
/
划分小节”,我可以将其分为几个小节,比如:
- 研究问题与框架 (h2或h3)
- 方法与模型 (h2或h3)
- 实验结果 (h2或h3)
但是规则要求“保留原文所有信息,不得增删改内容”。如果我加小标题,算不算增删改内容?
“不得增删改内容”通常指不改变正文的文字。加HTML标签是允许的,但加文本标题可能会被视为“增”。为了安全起见,最好不加自定义的标题文本,或者只用`
`分段。
或者,原文本身没有分段,我可以直接用一个`
`包裹整段,或者根据语义拆成几个`
`。
“每个自然段落用
包裹,长文用
`,代码用``,重要句子用``。 4. 保留原文所有信息,不得增删改内容。 5. 不要输出``/``/``等文档级标签。 原文内容: Detecting suspicious deviations in longitudinal medication trajectories requires distinguishing knowledge-consistent treatment variations from anomalous changes and linking local deviations to account-level decisions. We propose a knowledge-guided local-to-global (L2G) anomaly learning framework that addresses both challenges. It constructs proxy supervision at three granularities: medication occurrences, treatment episodes, and contiguous visit segments. Knowledge-consistent medication and episode replacements provide plausible negatives, while cross-category and cross-account perturbations provide suspicious proxies. A medication-multiset encoder and causal visit-sequence encoder feed a shared scoring head that produces visit-level anomaly logits. Top-\(α\) aggregation derives account-level scores from these logits, while account-level classification and class-balanced visit-localization objectives jointly optimize the same visit-level evidence. Experiments on two city cohorts compare L2G with native and proxy-adapted anomaly methods. On City-A, L2G outperforms the strongest comparators. Cross-city results show more consistent gains in localization than in account-level detection. 分析原文结构: 原文是一段连续的文本,没有明显的列表、引用或代码。可以将其划分为几个逻辑段落,或者直接用一个``包裹,或者用`
`/`
`划分小节。 为了符合“长文用
/
划分小节”,我可以将其分为几个小节,比如: - 研究问题与框架 (h2或h3) - 方法与模型 (h2或h3) - 实验结果 (h2或h3) 但是规则要求“保留原文所有信息,不得增删改内容”。如果我加小标题,算不算增删改内容? “不得增删改内容”通常指不改变正文的文字。加HTML标签是允许的,但加文本标题可能会被视为“增”。为了安全起见,最好不加自定义的标题文本,或者只用`
`分段。 或者,原文本身没有分段,我可以直接用一个`
`包裹整段,或者根据语义拆成几个`
`。 “每个自然段落用
包裹,长文用
/
划分小节” -> 如果不加标题文本,怎么用h2/h3划分?如果不加文本,h2/h3里没内容不行。所以可能不需要强行加h2/h3,因为原文不长,且没有明显的章节划分。直接分成几个`
`即可。 将原文拆分成几个自然段: 段1:Detecting suspicious deviations in longitudinal medication trajectories requires distinguishing knowledge-consistent treatment variations from anomalous changes and linking local deviations to account-level decisions. We propose a knowledge-guided local-to-global (L2G) anomaly learning framework that addresses both challenges. 段2:It constructs proxy supervision at three granularities: medication occurrences, treatment episodes, and contiguous visit segments. Knowledge-consistent medication and episode replacements provide plausible negatives, while cross-category and cross-account perturbations provide suspicious proxies. 段3:A medication-multiset encoder and causal visit-sequence encoder feed a shared scoring head that produces visit-level anomaly logits. Top-\(α\) aggregation derives account-level scores from these logits, while account-level classification and class-balanced visit-localization objectives jointly optimize the same visit-level evidence. 段4:Experiments on two city cohorts compare L2G with native and proxy-adapted anomaly methods. On City-A, L2G outperforms the strongest comparators. Cross-city results show more consistent gains in localization than in account-level detection. 或者,把“知识引导的局部到全局(L2G)异常学习框架”相关的用``强调一下? 规则说“重要句子可用”。 比如:
Detecting suspicious deviations in longitudinal medication trajectories requires distinguishing knowledge-consistent treatment variations from anomalous changes and linking local deviations to account-level decisions. We propose a knowledge-guided local-to-global (L2G) anomaly learning framework that addresses both challenges.
It constructs proxy supervision at three granularities: medication occurrences, treatment episodes, and contiguous visit segments. Knowledge-consistent medication and episode replacements provide plausible negatives, while cross-category and cross-account perturbations provide suspicious proxies.
A medication-multiset encoder and causal visit-sequence encoder feed a shared scoring head that produces visit-level anomaly logits. Top-\(α\) aggregation derives account-level scores from these logits, while account-level classification and class-balanced visit-localization objectives jointly optimize the same visit-level evidence.
Experiments on two city cohorts compare L2G with native and proxy-adapted anomaly methods. On City-A, L2G outperforms the strongest comparators. Cross-city results show more consistent gains in localization than in account-level detection.
检查是否保留了原文所有信息,没有增删改。 原文: Detecting suspicious deviations in longitudinal medication trajectories requires distinguishing knowledge-consistent treatment variations from anomalous changes and linking local deviations to account-level decisions. We propose a knowledge-guided local-to-global (L2G) anomaly learning framework that addresses both challenges. It constructs proxy supervision at three granularities: medication occurrences, treatment episodes, and contiguous visit segments. Knowledge-consistent medication and episode replacements provide plausible negatives, while cross-category and cross-account perturbations provide suspicious proxies. A medication-multiset encoder and causal visit-sequence encoder feed a shared scoring head that produces visit-level anomaly logits. Top-\(α\) aggregation derives account-level scores from these logits, while account-level classification and class-balanced visit-localization objectives jointly optimize the same visit-level evidence. Experiments on two city cohorts compare L2G with native and proxy-adapted anomaly methods. On City-A, L2G outperforms the strongest comparators. Cross-city results show more consistent gains in localization than in account-level detection. 拆分后的文本完全一致。 关于 `Top-\(α\)`,在HTML中,反斜杠和括号可以直接保留,或者不需要转义,因为 `\` 和 `(` `)` 在HTML文本中是安全的。直接输出 `Top-\(α\)`。 是否需要用 `` 或 `
`? 如果不用,也符合“长文用...”,因为这篇不算长文,只是一段摘要。如果非要加,可以加没有文本的标题?不行。所以不加标题是合理的。 再检查规则: 1. 只输出HTML片段本身,不要Markdown围栏、不要解释; -> 满足 2. 每个自然段落用
包裹,长文用
/
划分小节; -> 满足(用p包裹) 3. 列表内容用
- ,引用用
,代码用
,重要句子可用; -> 满足(用了strong)
4. 保留原文所有信息,不得增删改内容; -> 满足
5. 不要输出//等文档级标签。 -> 满足
最终HTML:
Detecting suspicious deviations in longitudinal medication trajectories requires distinguishing knowledge-consistent treatment variations from anomalous changes and linking local deviations to account-level decisions. We propose a knowledge-guided local-to-global (L2G) anomaly learning framework that addresses both challenges.
<
,代码用,重要句子可用; -> 满足(用了strong) 4. 保留原文所有信息,不得增删改内容; -> 满足 5. 不要输出//等文档级标签。 -> 满足 最终HTML:Detecting suspicious deviations in longitudinal medication trajectories requires distinguishing knowledge-consistent treatment variations from anomalous changes and linking local deviations to account-level decisions. We propose a knowledge-guided local-to-global (L2G) anomaly learning framework that addresses both challenges.
<