There were 2 rounds both of which was conducted on Google Meet platform. Initially document verification was done online and then we were alloted to different Panel. There were 3 panels in round 1. The panel which I got consisted of 3 professors. They were Prof. Manish Singh(P1), Prof. Srijith P.K.(P2), and Prof. Maunendra Dasekar(P3).
Round 1 was of 10-15 mins. Only 2 of the 3 professors asked question to me.
P1 - In which dept are u pursuing your B.Tech. Then he asked of I knew the concept of random variables.
Me - Yes Sir
P1 - He shared a google docs link in the chat window and there one question was given. It was:-
X is a random variable which can take values 0,1,2. Given E(X)=1.2. Find the maximum value of E(X^2)
Me - I had to type out the solution and explain each and every line which I have written and how was I approaching the problem. In the end I was able to answer it.
P1:- Are u comfortable with DS and Algorithm.
Me - Yes sir.
P1:- How can we find the kth min from an array of n numbers.
Me - We can sort the array and get the value of kth index.
P1 - Time complexity of the code
Me - O(nlgn)
P1 - any better solution
Me - After thinking for a while i said we can use the concept of partition in quick sort.
P1 - Time complexity
Me - I analysed and got something like O(n^2). (But after the exam I searched and on geeksforgeeks site it was written this approch take O(n) time if implemented.)
P1 - Is there any other way as well which is better.
Me - I could not think of any better way. (There is one more approach using heap which is better).
P3 - Gave a list of numbers in the docs and asked me to find the median of the list of numbers given.
Me - Gave the answer and discussed the two formula for even no. of element case and odd no. of element case in the list.
Round 1 was over and we were said that if selected we would be sent email for 2nd round which would take place around 3pm. I received an email stating I was selected for the round 2 interview.
Round 2 there were 2 panel named panel 4 and panel 5. I was alloted to panel 4 where there were 4 professors namely Prof. Abhinav Kumar(P1), Prof. Vineeth N Balasubramanian(P2), Prof. SaiDhiraj Amuru(P3), and one more professor(P4) whose name I do not remember. Only 2 out of the 4 prof. asked question to me. Round 2 lasted for 20-25 mins.
P1 - Introduce yourself. And why do u want to pursue M. Tech.
Me - Answered
P1 - Are u comfortable with probability.
Me - Yes Sir.
P1 - Do u know about gaussian random variable.
Me - No sir.
P1 - What random variables do u know.
Me - Binomial, Poisson, Uniform.
P1 - One basic question on conditional probability related to bags and balls without replacement.
Me - Answered that question.
P1 - You said u knew Uniform Random variable. Then given X is a uniform random variable from 0-2 and Y is a uniform Random variable from 1-3. Z ks another random variable which is defined as Z=X+Y. What is the PDF for Z and the graph using which it is represented.
Me - Z will be defined from 0-3 and the graph will be parallel to X axis from 0-3.
P2 - What have u studied in ML like which concept in ML would u be liked u asked about.
Me - Regression.
P2- Do u know Logistic regression.
Me - Not so much clear with the concepts.
P2 - Okay no problem. Do u know linear regression.
Me - Yes sir
P2 - Then tell me which concept we should follow if we want our regression line to be a curve instead to a straight line
Me - Multiple Linear Regression.
P2 - Any other algorithm which u would be liked to ask question about.
Me - Knn classifier
P2 gave me one situation and asked me what should be the value of k bigger, smaller or in between the total range of data.
Answered it.
P2 - Do u know about basis of a vector.
Me - No sir
And the interview was over and then the next day we were sent a list of projects in which we have to fill in our preferences for the project.
For the future aspirants I would recommend everyone to go through probability and linear algebra(especially vectors part) thoroughly and also some basic concept of ML algorithms which u can get by doing some courses online.
