1,584 views

Hi guys ,this is Shounak ( worst_engineer in GateOverflow ) .

So , finally after 2 GATE exams , I am able to use my GATE score to get admission in my dream place ... IIT.... IIT Hyderabad to be precise. It is MTech (RA) (Cut-Off here).

My research area in ML and Data Analytics.

I can not thank Arjun sir and Kathleen madam enough for GATE preparation help that he did through this portal.

As it was RA , I had to appear for written test and interview.
Now , coming to the interview. My work experience in Oracle had helped me.. specially for database concepts and programming skils . Also , I was studying some courses in Data Science ( Andrew Ng's flagship course ) , that also helped to some extent.

Questions asked me to were  like these :

1) Why do you want to come to MTech ?
- told about my intention to come to research in Data Analytics and Data Science.

2) You have given Data Analytics is your only preference. Are you aware that if you don't clear the interview , you won't get offer ?
- told yes.
Now , comes technical questions

3) Given a min heap , how will you find 3rd (or kth) smallest element ?
- I gave the approach. Remove root , then call Build heap which is O(n) .. continue for 3 times .They told me if I can optimize. I asked whether I am allowed to use extra space. I gave a solution where I will take an empty array. I will remove each element and put its children in array. But , instead of calling Build heap for entire heap, I will do Build heap and heapify for k elements. Which will be O(klogk). they were somewhat satisfied.

4) What is TLB
- told.

5) How would you find mean and variance of number of sixes hit in an test match inning ?
- I gave binomial distribution. But , professor gave a hint that , number of balls in an inning of test match is large as compared to number of sixes being hit. So, I told normal distribution. Converted binomial expression to normal variate.

6) How do you sort a very huge array ( may be 1 TB size) parallely ?
I gave solution that I will divide it in some n-threads and I will divide the array and send the index numbers to these threads.
Professor was asking for more details.. I was not able to convince entirely. Specially about merge procedure.

After that some more questions were there , which I forgotten.

After this round , there was lunch break.

After lunch break, I had interview with professors in Data Analytics field.

Questions were like this

1) Introduce yourself.
2) Matrix multiplication and graph theroy based questions.
3) Mean ,mode, median and which one to use in case of outliers in data.
4) Eigen value and eigen vector
5) They asked me , if I know anything on ML.
I told linear regression.
They asked me explain in brief.

I explained RMSE , TSS  and coefficients etc in details.

After that interview was over .

And after 10 days I got the offer letter :)

I would like to thank Arjun sir , Praveen sir , Bikram sir , Debashish , Digvijay , Pooja , Aakash and numerous other talented folks. Please forgive me , if I forget to mention anyone's name here.

posted Feb 5 in 2017
edited Feb 5 | 1,584 views