Unverified50% confidenceFactExact time
SVM是判别模型，其原始输出是样本到决策超平面的有符号距离，并非概率
1
Sources
50%
Confidence
Long-term
Relevance
7/19/2026
First Seen
Sources
Related Claims
Unverified核技巧使SVM可以在不显式计算高维映射的情况下实现非线性分类79% similarUnverifiedSVM的求解过程本质上是一个带约束的二次规划问题，计算复杂度主要由样本数n决定，而非特征维度d76% similarUnverifiedSVM的对偶形式通过引入拉格朗日乘子并利用KKT条件转化，其计算复杂度只依赖于样本数而非特征维度71% similarUnverifiedRepresentative algorithms in machine learning include Support Vector Machines (SVM), decision trees, and Naive Bayes.67% similarUnverified奇异值分解（SVD）是推荐系统中协同过滤算法的数学核心63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/558796API
curl https://kongchang.com/api/v1/knowledge/claims/558796MCP
get_claim(id=558796)