Unverified50% confidenceFactExact time

Training/serving skew can be further subdivided into three types: data skew, feature transformation skew, and algorithm skew.

1

Sources

50%

Confidence

Long-term

Relevance

9/1/2026

First Seen

Sources

Production-Grade ML System Design: Nine Engineering Challenges Beyond the Model

Related Entities

Training/Serving Skewconcept

Related Claims

Unverified数据增强从偏差-方差权衡视角看，本质是通过增加训练样本的有效多样性来降低模型方差72% similar Unverified参数量相似的情况下，训练数据质量分布的差异往往比架构差异更能决定特定场景的表现上限70% similar Unverified训练-部署鸿沟指模型因训练与部署环境的数据分布偏移、软件依赖版本差异或硬件指令集不同而导致的性能下降69% similar Unverified不同模型系列的工具调用输出格式差异根源于指令微调阶段使用的训练数据差异68% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/839427

API

curl https://kongchang.com/api/v1/knowledge/claims/839427

MCP

get_claim(id=839427)