This was more difficult than any high-tier big-tech interview that I've ever had. Question 1 was something that a DS would never do, I can't remember ever seeing somebody implement a model in pure numpy other than in a college course maybe where you're learning about the algo itself. This was something that realistically might take a couple days to "do it right" and a quick version of this would be about as quick and dirty as possible. Question 2 had the level of complexity and the amount of different tasks that was easily on par with every take-home DS assessment I've had where I've been emailed a csv and a list of questions and given a number of days to solve it using the tools I want to, in a very open-ended manner, with the ability to email the company with any clarifying questions and google anything I want. What I think was absurd about this was not the questions themselves, I think in some cases they were good questions, but rather the fact that they put them on a platform like hackerrank with a pretty unrealistic time limit. Everything is expected to be very well documented, not just how or why it works but "I did this because, this is what I saw, these are the implications etc." Talk about model results and all of your logic, and make visualizations related to your results.
LINEAR DIFF EQUATION SYSTEMS WIKIPEDIA FULL
This was very much a full on build a model, feed it some data, talk about the data a bit, etc.Ģ.) A machine learning problem where you have to do a bunch of data exploration and visualization, build and tune a model in a heavily time-limited test where your code is being run on some dinky VM. This was an algorithm that nobody would ever build from scratch in a real-world role. The hackerrank test had 3 sections and was only a few hours long:ġ.) A question where we had to build a simple and commonly used algorithm, but from scratch using only numpy. Sorry I know this sounds a little vent-y, pretty mad.
At this point I have a few different jobs under my belt and a few years of experience, I've done a number of data science interviews, I've had some truly absurd ones but the one I just had left me dumbfounded, and I'm curious about other people's experience.Īlso, I'm curious about what people think of my experience, if I'm being too critical or unrealistic etc. I just finished a hackerrank test for a position at a barely mid-tier company. I know these are all very broad questions but if anyone has any advice/experience in the things I wrote about please share! (I would plan to work for 2-3 years before moving back to the states) 2) will I be a desirable hire? Yes I have a big 10 degree but with no US work experience, will that automatically make me a risky hire? I wonder if it’s better to work in the states first then go abroad and then come back again :/ I don’t know anyone that’s done big career moves like this so I can’t tell if I’m doing wishful thinking or just being plain stupid. I figured maybe it would be worth my while to move to Europe to work/travel and then come back to the states to really start a life.Ĭoncerns I have is 1) I know a job in America would be more “prestigious” and pays more but as a early 20 something, my naive brain is telling me to take advantage of my youth instead of doing it later/regretting it later if I happen to not ever make the move. Everyone around me has been getting rejected for internships and jobs. This may be a shot in the dark but is there anyone that worked in Europe right out of college and moved back to the states after a few years? I’m about to graduate and I’m not necessarily the cream of the crop but I’m also not the bottom in terms of skill.
0 kommentar(er)
