Unverified50% confidenceFactExact time
sklearn的SVC使用John Platt于1999年提出的Platt Scaling方法，在SVM训练完成后额外拟合一个Sigmoid函数将决策值映射到[0,1]区间来输出概率
1
Sources
50%
Confidence
Long-term
Relevance
7/19/2026
First Seen
Sources
Related Claims
UnverifiedGridSearchCV将参数网格的穷举遍历与k折交叉验证结合，导致训练次数乘以k倍65% similarUnverified强行取极值样本会引入离群点，迫使SVM构建更复杂的决策边界，高阶poly核在这种数据上收敛更慢65% similarUnverifiedPlatt Scaling通过Sigmoid函数P(y=1|x)=1/(1+exp(Af(x)+B))映射概率，参数A和B通过最大似然估计在留出集上拟合62% similarUnverifiedLoRA leverages the low-rank approximation principle of SVD to decompose the incremental updates of large weight matrices into the product of two low-rank matrices, thereby compressing the number of trainable parameters from billions to the millions range.61% similarUnverified支持向量机（SVM）由Vapnik和Cortes在1995年提出，其核心是正负类之间决策边界的间隔最大化59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/558030API
curl https://kongchang.com/api/v1/knowledge/claims/558030MCP
get_claim(id=558030)