ML 常见问题总结

分享一些自己总结的ML面试常见问题。自己遇到的问题基本都在里涵盖了。希望能帮到大家。

祝大家面试好运，都能拿到想要的offer！！

求米~急需大米看面经。如果觉得有用记得赏米哦~多谢！

1.        How to check overfitting? How to deal with it?
2.        Could you design a query embedding for Amazon teams?
.google  и3.        Ensemble algorithm? (Random forest; feature and data replacement; reduce variance
4.        How to split a tree?
5.        What metrics would you use in a classification problem?
6.        How to deal with an imbalanced data set
7.        What loss function will you use to measure multi-label problems
8.        Let’s say now you want to identify a threshold for a classifier that predicts whether a customer will sign up to prime or not. What criteria could we use to find the threshold?
9.        In the model we developed, we have a billion positive samples and 200,000 negative samples. If you were to review our model before we put it on the website, what would you look for in the model to ensure this model is not bad?
10.        Could you describe how you train this Context-awareness entity ranking model?
11.        1. 解释regularization，为什么要regularization， l1和l2的区别
12.        2. 解释bagging和boosting， boosting要求每个iteration的training set都一样吗？（it depends）
13.        3. 说一些dimensions reduction的方法（我说了pca），

您好！
本帖隐藏的内容需要积分高于 188 才可浏览
您当前积分为 0。
使用VIP即刻解锁阅读权限或查看其他获取积分的方式

游客，您好！
本帖隐藏的内容需要积分高于 188 才可浏览
您当前积分为 0。
VIP即刻解锁阅读权限 或 查看其他获取积分的方式
Unlock interview details and practice with AI
Curated Interview Questions from Top Companies

ions of linear regression
37.        7.How to deal with categorical features
38.        8. formular of logistic regression and the loss fun
39.        10. Explain how Kmeans is working
40.        11. If the linear regression dataset is duplicated, how would all the parameter change in your model?
41.        12. In linear regression, how will multicollinearity impact the cofficients and variance?
42.        13. If the correlated features should be excluded in random forest model?
43.        14. what is  bias-variance trade-off
44.        15. how to deal with overfitting
45.        16. how to deal with imbalanced dataset
46.        17. How to deal with missing data

很多问题都是罗列其他楼主的，感谢大家的贡献！！！

感谢楼主分享

感谢LZ分享~

感谢LZ分享~

只有100积分，还不够188。看不到中间的21到36的问题，可以求好心人分享下吗

谢谢楼主分享！

感谢楼主分享

屁特里 发表于 2021-10-6 20:08
只有100积分，还不够188。看不到中间的21到36的问题，可以求好心人分享下吗

我也是...拜托可以分享一下吗 ==

非常好的总结！！！感谢lz分享