Project github address: bitcarmanlee easy-algorithm-interview-and-practice
welcome everyone to star, leave a message, and learn and progress together
The interview process of a TMD factory is recorded as follows
first round
1. Introduce the project (25min)
2. Talk about fm
3. What methods did you use for the recall? How to construct the training set of collaborative filtering/ucf icf? (10min)
4. Talk about word2vec
5. Add, delete, and modify the two strings of word2vec . How many times does a become B at least?
second round
1. Project description (20min)
2, root out the details of the project
3, speaking at the difference deepfm and fm
4, speaking at xgboost
5, speaking under the tree
6, how to analyze data
(in fact, asked more than these, we have the impression These ones)
Follow-up
First, the department head introduces himself
,
then my self-introduction, and finally introduces his project team and work content to me