Describe the difference between L1 and L2 regularization, specifically in regards to the difference in their impact on the model training process itself.

Describe the difference between L1 and L2 regularization, specifically in regards to the difference in their impact on the model training process itself.

题目类型: 技术面试题

这是一道技术面试题，常见于澳洲IT公司面试中。

难度: medium

分类: problem-solving

标签: Apple Data Science

目标岗位: Data Scientist

目标公司: Apple

参考答案摘要

回答要点：L1 adds |w| promotes sparsity/feature selection; L2 adds w^2 shrinks weights smoothly. Training-wise, L1 can yield non-differentiability at 0 and more sparse solutions; L2 is stable and distribut...

本题提供 STAR 原则详细解答和技术解析，登录匠人学院学习中心即可查看完整答案。