Unverified50% confidenceDecision RuleExact time
When dealing with highly imbalanced classification datasets, one should consider SMOTE oversampling or shifting the evaluation metric from accuracy to F1 score.
1
Sources
50%
Confidence
Long-term
Relevance
9/7/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn class-imbalanced scenarios, the F1 score better reflects the model's ability to recognize the minority class than accuracy does.75% similarUnverifiedDuring fine-tuning, hard negatives (e.g., positions in the same field but at a different seniority level) should be used instead of random negatives to obtain maximum gains.71% similarVerified若用含测试集的全量数据计算均值和标准差，会导致预处理泄漏，使评估指标虚高，无法真实反映模型泛化能力69% similarUnverified对MoE模型的量化需要格外谨慎，因稀疏激活使专家参数数值分布集中，低比特量化易造成路由偏差导致能力退化68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/869690API
curl https://kongchang.com/api/v1/knowledge/claims/869690MCP
get_claim(id=869690)