Unverified50% confidenceFactTime unknown
从贝叶斯视角理解，L1正则化对应参数的拉普拉斯先验，L2正则化对应高斯先验
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026
Sources
Python机器学习实战：算法推导到代码复现全流程教程
bilibiliPython课程视频
Related Claims
Unverified在满足约束等距性（RIP）的字典下，L1松弛（LASSO）可以精确恢复稀疏解，为L1正则化提供理论保证68% similarUnverified选择性量化思路与LLM领域的GPTQ、AWQ等量化方法一脉相承67% similarUnverified当特征向量被L2归一化后，余弦相似度等价于点积运算67% similarUnverified伯努利分布属于指数族，其自然参数与均值参数之间的映射恰好是sigmoid函数，逻辑回归是GLM框架下伯努利分布的规范链接函数选择65% similarUnverified卡尔曼滤波是线性高斯假设下的贝叶斯最优估计器，其预测-更新两步迭代对应先验与后验概率的递推计算63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/60910API
curl https://kongchang.com/api/v1/knowledge/claims/60910MCP
get_claim(id=60910)