Unverified50% confidenceFactExact time
AdaBoost中弱学习器权重公式为α_t = (1/2)ln((1-ε_t)/ε_t)，其中ε_t是该轮的加权错误率
1
Sources
50%
Confidence
Long-term
Relevance
10/5/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAdaBoost中正确分类样本权重乘以e^(-α_t)，错误分类样本权重乘以e^(+α_t)，然后归一化78% similarUnverifiedAdaBoost的权重更新公式来源于对指数损失函数L=Σ_i e^(-y_i F(x_i))的逐步最小化推导71% similarUnverified工作中引入的权重衰减常数λ*定义为 λ* = 1 - αₗᵣ·λ，其中αₗᵣ为学习率，λ为原始权重衰减系数70% similarUnverified论文证明学习得到的生成分布在Wasserstein-p距离下满足有限样本误差界W_p ≲ n^(-1/d) + n^(-1/(2p))(log(1/ξ))^(1/(2p))67% similarUnverified在真实映射为均匀随机的假设下，论文证明了错误率下界为 E ≥ (M/N)δ*(B/M) + (1-M/N)(1-1/K)65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/977407API
curl https://kongchang.com/api/v1/knowledge/claims/977407MCP
get_claim(id=977407)