The AI department of IIT Hyderabad had shortlisted the candidates based on gate score and academic performance. They had selected around 40 candidates for the interview. Due to covid pandemic the interview was organized on Google meet. The interview was scheduled on 17 July 2020.The candidates had to be online at 10 am for document verification.

Documents were verified and I had to join the interview using a different Google meet link sent.

They were 3 professors in the interview panel.

Prof 1: Give a brief introduction about yourself.

Me: Answered.

Prof 1: So, which are your subjects of interest? Do you have a liking for probability and Linear Algebra?

Me: I told yes So, they started asking questions on probability and linear algebra.

Prof 2: What are independent events?

Me: Answered

Prof 2: How do you identify independent events? Formula for independent events & Bayes Theorem.

Me: Answered

Prof 2: Draw and show in Venn diagram 2 events A and B which are independent.

Me: Drew the Venn diagram.

Prof 2: So, from your diagram you mean to say independent events cannot be identified just by diagram?

Me: Yes, Sir We need to know the probability of intersection of A and B.

Prof 2: Create your own example for Independent event and explain how you will conceive that they are independent.

Me: Answered with respect to rolling of dices.

Prof 3: Have you taken up Linear Algebra in UG?

Me: Yes

Prof 3: Ok you are given a function and told to find minimum how will you find?

Me: Answered

Prof 3: In continuation to the question find the minimum of f(x)…

Me: Differentiated and told the points of minima.

Prof 3: Gave 2 vectors A and B and said find a 2*2 vector X such that AX=B

Me: Gave an answered directly and trial and error

Prof 3: I know you might have tried the values now what if it is 100*100 matrix.

Me: Understood and got the answer. The question was based on Eigen Vectors.

Prof 3: Asked a few questions on Eigen vectors.

The interview took around 25 minutes. Later on, the same day, I received a mail for round 2 of their interview. They had shortlisted 14 candidates.

Interview panel had 3 members. One of the professors was the same as round 1.

Prof 1: What are your UG projects?

Me: Answered

Prof 1: Your projects are based on Cryptography and Security, then why AI?

Me: Explained

Prof 1: What all do you know about Machine Learning?

Me: Answered

Prof 1: What are classifiers? Tell me some binary classifiers.

Me: Answered

Prof 1: What if I have multiple classes. You have studied multi class classifiers?

Me: Bit confused

 Prof 1: How can you make binary classifier to work as multi-class classifier ?

Me: Answered

Prof 1: Given 1 vs 1 and 1 vs all approach if there are 10 classes how many classifiers are generated?

Me: Answered

Prof 2: What is Logistic Regression. Why is it called Regression?

Me: Answered.

Prof 2: You said you know about KNN classifier. State the disadvantages of using it.

Me: Answered not quiet satisfied.

Prof 3: Given a double differentiation of the function is always positive. What can you infer?

Me: Answered

Prof 3: What are convex functions?

Me: Answered

Prof 3: Can you give me mathematically how logistic regression error function is derived?
Me: Not Answered

 

Prof 3: Thank you, we are done

 

The round 2 of interview was basically to check if we have prior knowledge about Machine Learning, though professors told that its ok to start from zero if anyone has not undergone ML during UG courses. The professors were very helping and cooperative. The interview was a great learning experience to me.
posted in Interview Experience Aug 9, 2020
3,263 views
1
Like
0
Love
0
Haha
0
Wow
0
Angry
0
Sad

2 Comments

2 Comments

Like
1
Can you tell your GS and CGPA?
Like

Hi, Thank you very much for uploading your experience.

Could you explain the conversation from the screenshot below? Did you mean to write a "2*1" vector X. How is this related to Eigen Vectors. As per my knowledge, they help when we are trying to multiply a high power of A with x. It would be really helpful if you could give an example.