Kim Question check (https://www.geeksforgeeks.org/samsung-interview-experience-set-30-campus/)
They actually gave a very big question. The challenge in the question is that you should be able to analyse the complete question in very limited time. Interviewers were chill and they were giving hints. I answered the brute force solution and Optimised it based on interviewer’s hints. Of course they asked us to write the pseudocode but you should also be able to convince interviewer with your answer.
Technical Round 1 and 2 (I don’t remember which question is asked in which round):
1.Tell me about yourself (Try highlighting the points mainly which are related to the profile in this case mainly your interest in ML and if possible in related areas like DL and NLP also speak about your achievements)
2.In both the rounds they asked me to explain my second year intern project, I actually published a conference paper (this played an imp role while selection) based on this work so they were interested in this project. Your project may be the best but convincing the interviewer is the key!
3.A simple question on Bayes Theorem, if you know Bayes theorem you will be able to answer it.
4.Difference between generative and discriminative classifiers and example for each.
5.Linear regression and Logistic regression everything from hypothesis to the Loss function. You are expected to be known of the advantages of logistic regression over Linear Regression.
6.Do you know CNN’s? Yes I have some idea but never worked on them!!
I explained him why does it is called neural network he then asked me why it is called as convolution, I was thinking something in terms of normal convolution (but din’t explained him may be you should!) but never give up until interviewer proceeds to next question on his own.
7.Do you know Softmax regression and LSTM? Explained about Softmax only.
8.You have infinite denominations of 1, 5, 10 how many ways can you make 605 sum. He asked this as a puzzle but thanks to dynamic programming I answered it as a coding question he was a bit ok with this.
- What are the conditions on dataset for using Linear Regression. I answered some algorithms expect the dataset to have Gaussian noise but not sure of linear regression in particular.
- They asked me why your projects are pure ML based but not on Deep Learning and NLP (They use DL and NLP in their projects), I said I never got an opportunity to work on those things but willing to work on them.
11.What is Likelihood and probability, can we use them Interchangeable?
12.What is entropy? Explain it’s physical significance, Give the formula for this explain intuition behind this formula.
13.Some basic question on Parazen window.
I would recommend you to have good understanding of all the important concepts of machine learning, not only what an algorithm does but in what cases will it be preferred and how the algorithm is formulated (like some algo’s do least square error minimisation and some use maximum likelihood concept).
There is an online course on machine learning by Caltech university (Prof Yasser Abu Mustafa) in YouTube, there the concepts are explained in an incredible way. Thanks to Caltech and Prof.Yasser.
Also most of the concepts are explained in a very intuitive way in blogs like Medium and Towards Data Science. I would highly recommend to have a look at these.
MATLAB has an article on choose classifier it will help you in getting some overall idea about ML.
Hacker earth also has some articles related to ML and Image Processing they also will help you in getting some good feel about ML.
Asked about a project.
Gave a very long and clumsy question but if analysed properly it turned out to be a basic directly or inversely proportion based question.
Basic HR questions but prepare for them too, Some questions being
- Why Samsung? (Show interest in the company it is very imp).
- What are general problems you face working in a group?
- Are you a leader or a follower?
- Many of your projects are mostly mathematical are you ok with working on DL and Neural Nets based projects.
Overall 3 people got selected 2 for research profile (I am one among them) and 1 for developer.
Thanks a lot to Geeks for helping me mainly in the coding part.
Attention reader! Don’t stop learning now. Get hold of all the important DSA concepts with the DSA Self Paced Course at a student-friendly price and become industry ready. To complete your preparation from learning a language to DS Algo and many more, please refer Complete Interview Preparation Course. In case you are prepared, test your skills using TCS, Wipro, Amazon and Microsoft Test Serieses.