Deloitte Data Science interview questions

2 rounds of interview were conducted. 1st was a technical interview conducted by a Data Scientist of similar profile as mine. It was a good interaction, and the interviewer was clearly well trained in conducted tech interviews. 2nd was supposed to be a "behavioral" interview conducted by a "manager" position employee who was extremely unprofessional. He had no knowledge how Data Scientists operate and no understanding of machine learning, which is al-right since he is not supposed to be an expert of sciency stuff and was asked to conduct a behavioral interview. But he kept on asking silly questions about data science. Explaining data science projects to someone who is not familiar with the subject is something that we as data scientist should be good at, by clearly articulating using laymen terms. But the person must be aware that he is not familiar with things and should listen to my explanations properly. However explaining data science to someone who thinks he understand it but obviously had no idea what I am talking about should not be the judge of my skillsets. 2nd interviewer was highly biased, untrained, and unqualified to conduct a technical questioning & he wasn’t suppose to. I don’t know if this was properly communicated to him by HR or not or maybe he had no experience taking interviews at all.

Applied online and about 2 weeks later got a HireVue interview. They will ask some behavioural questions like why Deloitte, why consulting, describe your previous experience/education, greatest strengths/weaknesses. The HireVue also has a coding portion where you have to solve 4 challenges that are similar to leet code easy. Other people have posted some of the questions under Data Analyst/Analytics interviews. Some questions were: 1. Calculate mod without using modulus 2. Get the Fibonacci number 3. Parse, join and output 2 CSV like strings 4. Given a string of moves, will you return to the starting point About a month after the HireVue got invited for another round. This was a rapid fire statistics/machine learning theory round where they will ask any questions in the domain of math/statistics/machine learning and not necessarily related to your previous experience. They will start by asking easy-hard questions on a broad range of topics and gradually adjust the difficulty to match your knowledge. Some questions: What are the assumptions of linear regression. How to evaluate linear regression? What are some problems with R^2? How to check for multicollinearity? Discrete time vs continuous time markov chains? Different types of distributions and what can they be used to model? How does PCA work and what are some problems? What are assumptions of time series models? How do you deal with auto correlation? What is the bias variance trade off? Honestly thought I didn't get past that round but a couple days later they invited me for another round. This was more laid back and it was about past experiences and going through some case studies and thought experiments. They will also question decisions you made on past ML projects and ask slightly technical questions. It was more behavioral and to see how you problem solve. 2 weeks after that didn't hear back and I saw the job get reposted so assumed I didn't make the cut.