Tip 1 : Prepare resume well
Tip 2 : Get basics clear
Tip 3 : Keep practicing coding questions
Tip 1 : Keep it short and to the point, one page is more than enough
Tip 2 : Only keep things that you have actually done, don't fake it in your resume
The main intention of the round was to check my interest and my work. It was more of a screening round.
There were basic questions on my work, master's thesis and interest area.
What is SVM, draw and explain its loss function.
What is linear regression and how is it different from logistic regression?
theirs loss functions
What is entropy and gini index?
What are precision, recall, and f1 definitions? there was a numerical question related to them
What is RNN and why do we use them? Difference between vanilla RNN and GRUs



It is a well-known problem



Input: 'n' = 3, 'm' = 4
'queries' = [[1, 1], [1, 2], [2, 3]]
Output: [1, 1, 2]
Explanation:
Initially, the grid was:
0 0 0 0
0 0 0 0
0 0 0 0
After query (1, 1):
0 0 0 0
0 1 0 0
0 0 0 0
There is one island having land (1, 1).
After query (1, 2):
0 0 0 0
0 1 1 0
0 0 0 0
Since (1, 1) and (1, 2) share a common edge, they form one island. So there is one island having lands (1, 1) and (1, 2).
After query (2, 3):
0 0 0 0
0 1 1 0
0 0 0 1
Now there are two islands, one having lands (1, 1) and (1, 2), and another having land (2, 3).
So the number of islands after each query is [1, 1, 2].
It was extension to prev problem
Used BFS to solved the problem. It is a well-know problem too



It is a well-known problem



‘N’ = 3, ‘coins’ = {1, 2, 3}, ‘freq’ = {1, 1, 3}, ‘V’ = 6
For the given example, we can make six by using the following coins:
{1, 2, 3}
{3. 3}
Hence, the answer is 2.
It was an extension of the prev problem
It was more of Managerial + HR round
This round revolved around basic behavioral and situation-based questions.
Tip 1: There are no right or wrong answers
Tip 2: Think and then speak
Tip 3: Always show a positive attitude
What is transformers? and the loss function used in traning
What is positional embedding in BERT and its mathematical formula discussion
Explain multi-headed attention and its math
Why BERT give better results than RNN?
Explain the encoder and decoder architecture of BERT
When do we need high precision and when do we need high recall
Explain the AUC ROC curve
Explain CNN and it can be used with text.
Desing a question answering system for unstructured documents.
Tip 1: Ask clarifying questions
Tip 2: Always keep on discussing your approach

Here's your problem of the day
Solving this problem will increase your chance to get selected in this company
How do you remove whitespace from the start of a string?