What is the triple loss?
Applied Scientist Interview Questions
1,159 applied scientist interview questions shared by candidates
What is backpropagation in deep learning?
What is batch normalization? what are its limits? what other normalization could think about?
What is HOG? What is SIFT? How are they computed?
1) Tell me a time your project didn't turn out as expected. 2) LeetCode question requiring bfs/dfs or graph traversal. 3) Design an XXX system that does task YYY (note YMMV).
Based on Trees and Graphs, basic Leetcode stuff. Some other topics that are very useful to review : Greedy Algorithms/Dynamic Programming Recursion/Backtracking Graph Theory Linked Lists Most questions were moderately easy and directly from Leetcode. There are no onsite for the internship, just phone interviews. One of the interviewers is from the team, asks more questions related to the team you are interviewing for.
Signed an NDA
The interviewers are smart but super nice too. The questions aren't too difficult if your ML fundamentals are clear (incase you're unsure you'll find out yourself after the interview). Coding questions were easy-medium (harder side of medium) level.
Logistic regression, decision trees, deep learning
What methods do you use to evaluate that your models is trust-worthy?
Viewing 131 - 140 interview questions
See Interview Questions for Similar Jobs
Deep Learning ScientistMl EngineerData Scientist NlpSenior Machine Learning ScientistDeep Learning Data ScientistApplied Research ScientistNlp Data ScientistAi Machine Learning EngineerDeep Learning Research EngineerApplied ResearchJr Data ScientistArtificial Intelligence ScientistData Scientist TraineeDeep Learning / Machine Learning ResearcherJunior Machine Learning EngineerApplied Machine Learning EngineerSenior Machine Learning EngineerMachine Learning Researcher