Unverified50% confidenceFactExact time
Robust aggregation methods may misjudge legitimate updates from normal clients as anomalous under non-independent and identically distributed (non-IID) data, because differences in data distributions among clients themselves cause gradient directions to diverge.
1
Sources
50%
Confidence
Long-term
Relevance
10/30/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedNon-independent and identically distributed (non-IID) data is the most typical source of statistical heterogeneity in federated learning, causing client drift that manifests as convergence oscillation, accuracy degradation, and systematic bias against certain clients.77% similarUnverifiedData heterogeneity in federated settings is often called the Non-IID problem, i.e., the data distributions of different clients are inconsistent.77% similarUnverifiedSilent model degradation can be caused by a vendor updating the model, prompts being mistakenly changed, or a shift in the upstream data distribution, and the interface itself does not report an error69% similarUnverifiedWhen statistical heterogeneity exists in the data distributions of clients, the global model tends to cater to clients with larger data volumes and mainstream distributions, causing clients that deviate from the mainstream to exhibit significantly worse local accuracy.69% similarUnverified信息论中的数据处理不等式表明，经过任何确定性或随机变换后信息只会减少不会增加67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1113860API
curl https://kongchang.com/api/v1/knowledge/claims/1113860MCP
get_claim(id=1113860)