Message on Whatsapp 8879355057 for DSA(OA + Interview) + Fullstack Dev Training + 1-1 Personalized Mentoring to get 10+LPA Job
0 like 0 dislike
in Interview-Experiences by Expert (30,360 points) | 494 views

1 Answer

0 like 0 dislike

Microsoft Interview Experience for Machine Learning Internship | 2022 | ON CAMPUS


Round 1 (Written) :

62 mcq questions in 1 hour. The level of paper was tough. Only 7 were shortlisted. The questions were based on Machine learning Algorithms, probability and counting, and statistics. It was conducted on Mettl. If you are able to solve 50% of the part, then you can expect a call.


Round 2 (Technical Interview) :

Focus : Problem solving skills

The 1st question was based on the trees. I was given an input and a binary tree. My task is to delete all leaf nodes whose data is equal to the given input (wait it’s not that simple, you need to implement cascade delete in this question). I told him the approach and it was correct. Then he told me to write the code for it. I tried and reached very close to the result but missed one condition. Then he asked a few questions on neural networks, sentimental analysis. He gave me a problem about how to classify flight booking emails from other mails. I gave several ideas and approaches. It was an open ended question so it totally depends on the interviewer whether he likes your idea or not. After this round 1 was over. It lasts for 50 minutes.

Tip : Practice some questions on Trees and Graphs from gfg. Also read the puzzle section.


Round 3 (Face to face interview) :

Focus : Projects and Domain knowledge

He told me to introduce myself. Then he quickly jumped to my projects section and asked every project one by one in depth. He asked several things on Neural Network, CNN, Bagging Boosting, decision trees, activation layers. How to prevent overfitting, Regularization etc. How to mix more than one ML Algorithms. How to think, which algo should we apply in any particular scenario. That’s all about the second round. It lasts for 45 min.


Verdict : None of us got selected for this profile.

by Expert (30,360 points)