I applied through an employee referral. The process took 2 weeks. I interviewed at Capital One in Apr 2016
Interview
First is to talk your resume with recruiter for 20-30 mins,after that will be a hackerrank challenge.After that I have been scheduled a tech interview with data scientist, the process goes really good, I wonder why I didn't pass this round, they are very likely to ask modeling questions that they may currently stuck, if you are not able to give the answers they want to solve their current problem , they will reject you.
Interview questions [1]
Question 1
coins, string manipulation, define a logit function and input data to compute.
how to detect the Fraud with current data problem, you are expecting to answer how to deal with missing value, how to use classification method, which one is good to use?Later there will also be a problem which method is the least useful. What is the effect of FP and FN.How to change the model if the amount (one is large like $1,000,the other is $5) turned out to classified as the same.
I applied online. The process took 1 week. I interviewed at Capital One (New York, NY) in Apr 2016
Interview
applied online, got a OA on hackerrank. Four questions in two and half hours. The OA majorly aims to test your basic programming ability to read in text from standard input and then process them with basic data structure
Interview questions [1]
Question 1
implement a function to calculate matrix sum. calculate the probability of coin head
I applied online. The process took 4 weeks. I interviewed at Capital One (Richmond, VA) in Apr 2016
Interview
They first send you a dataset challenge which has a regression problem and an exploratory analysis problem. After that there's on site interview which has three rounds.
1.Behavioral interview, which had three questions. Nothing fancy, things about your experience with teamwork and your innovation etc.
2. Case interview, which was similar to the example they have on their website. Some simple arithmetic and some business logic. By this time I was still waiting to have at least one conversation about machine learning and data analysis, nothing.
3. Role play interview is just a technical interview in disguise. I think this is where I lost it. The woman who interviewed me was a statistician who had no appreciation for "modern machine learning models". She thought k-means was modern (ahem). Anyways, she started this ten minutes long rant about how linear regression is still the most robust of models and how modern machine learning research is ruining all that.
I didn't get the offer. But I don't think I'd have taken it if I was offered the job. They call themselves innovative, but they're still a bank! a little phony.
I asked a lot of questions about their work culture and it was clear that they still have all the qualities of a bank, so not a great place for computer engineers. Was a waste of time, really.
Interview questions [1]
Question 1
If you use random forest, how can you explain why it's working to your investors? Isn't linear regression the best?