A. Presentation on a technical topic of your choice (to a roomful of around 20 people, who will constantly ask you technical questions related to your presentation as you progress along) B. Whiteboard Questions (with different sets of interviewers): 1. Given a random number generator randN() that generates integers between 1 <= x <= N write a function that generates a number between 1 <= y <= M such that each integer in the range [1, M] has an equal probability of selection. Note that both N and M are known in advance and M > N. You can only use randN() and cannot use any other random number generating function calls. 2. Say there is only one kitchen in your office that is shared by your team and another team. Both teams have 15 members each. Every morning exactly 30 apples can be found in the kitchen that anybody in your team or the other team can take but of course is not required to (which means there could potentially be some apples leftover in the kitchen at the end of the day). You suspect that the other team is taking more than its "fair share" of apples and you want to test this hypothesis. But you are not allowed to ask individuals how many apples they ate, even if they are from your team and you cannot observe who is taking the apples in the kitchen. How would you go about evaluating your initial hypothesis? Write down the test that you would use. 3. You are given three stack data-strucures that ONLY support pop, push, and peek methods. One of the stacks contains 5 integers in arbitrary order. Write a program in your language of choice to create a sorted stack with lowest numbers at the bottom of the stack and (ideally) not using more than the 3 stacks provided. 4. You are presented with a rectangular surface (map-like), with friction and environmental data recorded at a limited number of arbitrary points in the surface. There are a number of key-points in the surface for which you need to calculate the friction value. [Note: These "key points" were spaced as if they were the points of intersection of grid-lines that sub-divided the total surface area into smaller squares]. How would you implement a KNN approach to estimate the friction values at the key points if you wanted a distributed/scalable solution. What is the ideal way to place the key points in the surface, i.e. how would you determine the distance/spacing between the key points? 5. Difference between l1/l2 penalty (lasso vs. ridge); Why use elastic net rather than simply using L1? What is the manner in which L1 penalty "drops" a subset of correlated variables aka how does it "choose" which correlated variable(s) to drop? 6. Given Sum[(x_hat - x_i)^2] for i = i, 2,...,n what value of x_hat will minimize the expression given here? 7. Same as (6) but the expression is Sum[abs(x_hat - x_i)]? Justify/explain your answer. 8. What are the different ways to regularize neural networks? 9. Say there is a function that is fitting three distinct models (i.e. independently of each other) with 14 million different time series', each of time-length 'l', and generating future forecasts (the input and forecasts are overwhelmingly positive values but some can be negative). Given that you have limited computing resources, say 200 cores, what is the best way to distribute the task of performing this operation. You can either use Spark to show how it can be done or you can simply describe the algorithm to get it done. They will ask you follow-up questions on the run-time complexity of your chosen algorithm(s). C. HR and Management Joint-Interview: These were largely generic HR/Management type questions. 10. Why this company and this position? 11. What do you see yourself doing in the next 5 years? 12. Describe a situation where you had disagreement with team members / colleagues and how you resolved it. 13. What is your ideal work environment? What kind of managemenet style do you like? 14. Have you been in leadership positions? What was your management style? 15. Did you face a situation where one of your team members was not motivated at all? And how did you resolve the situation. 16. Would you consider a management position in the future? 17. Describe a work situation where you felt a technical skill lacking in your part or that you really wished you had. 18. Why did you decide to go to [college/university name]? 19. Thinking about [project in your resume], what was a very challenging scenario that you faced and how did you handle it. 20. Other than your past projects that we have already discussed, what other projects are you especially proud of and why? 21. Do you already have offers from other companies? 22. How much time can we take to notify you of our decisions? 23. If you are hired, would you start straight away or would you need some time to make that decision? - Finally, at the end of the interview you will have an opportunity to ask them questions