Zendrive Campus Interview Experience for Associate Data Scientist

First Round:
There were 12 questions from Probability for 60 minutes. The questions were little tricky but solvable in the given time if the concepts are clear.

Second Round:

• Do you know how to deal with time series data?
• How do you find correlation between two variables?
• How do you handle overfitting?
• What is Regularization? What are the types of Regularization?
• What is Maximum likelihood estimation?
• What is a Bayesian classifier? What is prior and posterior probability? Does the prior probability change if we add more training data and why?
• There is a coin. We flip it 5 times and the result is HTTHT. How do you find if the coin is biased?
• There is a coin. We flip it 3 times and the result is HTH. The probability of H in first trail is p, in the second trail is 2p and in the third trail is 3p. Is the coin biased?
  What is likelihood?

Third Round:

• Given a data such that one of the classes is almost at the center like a circle and the other class is spread away from the center. What model would you use to classify?
• Given driver data. For each driver we have the distance he has traveled and the mistakes he as done in that travel. How would you rate the driver as best or worst?
• Given a swimming pool, which looks like a hemisphere of radius 5m. There is a dive board on the top of the pool and a swimmer jumps from the board into the pool. After he touches water, he goes 3m into the water and comes back. If he hits the ground of the pool, he gets hurt. What is the probability that the swimmer doesn’t get hurt?
• Three persons have 3 mobiles with GPS. Mobile A can predict location within 100m accuracy. Mobile B can predict the location within 500m accuracy. Mobile C can predict the location within 2Km accuracy. All the three people are stranded on a boat in the middle of the sea. Because of the currents, the boat changes its location continuously. At any point in time, we know the predictions from the three mobiles. The errors of the predictions are normally distributed. How do you find the exact location of the boat from these time series predictions?

Fourth Round:

• Given an n*n n?n maze with n^{2} n2 cells and given some doors between adjacent cells, find if there is a path from cell (1, 1) (1, 1) to (n, n) (n, n) and print the path.
• Given 10 vessels, with vessel 1 having 1 litre water and the remaining vessels are empty. You can perform only this operation: Pick any two vessels and make the water in both the vessels equal by transferring water from one vessel into another. What is the minimum value of water vessel 1 can contain after any number of operations.
• There is a parliament with n parliamentarians. Each parliamentarian can have enimity with atmost 3 parliamentarians. We want to divide the n parliamentarians into 2 groups such that a parliamentarian can have enimity with at most 1 other parliamentarian in his group. How would you divide the parliamentarians into 2 groups. What is your approach?
• Tell me why did you come to academics after doing job for 3 years? There will be lot of stress in ISI Kolkata. How are you handling it? Can you handle the stress after joining us?