Data scientist interview questions shared by candidates
Business Intelligence: Using the data you have compiled in the previous part answer the following BI QUESTIONS: - How Many Customers above 50 years old have taken up a loan? - How Many Females aged 30 to 40 have more than 2 products? (Female=0) - What is the average amount of Current Account(CA) Transactions for Male - Who Had a previous Loans - How many females did not have a previous loans and who are aged: <20, 21-40, 40+ This code needs to be provided 12 hours ahead of the Round 2 interview. Predictive Modelling: The end goal is to create a model allowing you to identify customers more likely to take on a loan. We have kept an hold out sample, the quality of your work will be partly based on the hit rate you have been able to achieve based on your prediction. You will need to build the Test sample using the TEST_*.CSV files. The only thing missing is the loan flag. You will still be able to score those 2000 Individuals and put them into 5 groups • Very High Likelihood • High Likelihood • Medium Likelihood • Low Likelihood • Very Low Likelihood For each group define what you expect the % of Loan subscribers will be. You will need to define optimum side of High and Very High likelihood group based on business relevance. This code needs to be provided 12 hours ahead of the Round 2 interview. Data Scraping Linkedin Test: Using the provided dataset, develop a code which will allow the following: Search for each individuals on Linkedin and retrieve • Current Job • Job Title • Company name • Education • Number of connections The code must automatically populate the dataset, limited manual intervention is allowed but you must demonstrate that the code is scalable. You can use any programing Platform of your choice (Python, R, ect…) You must flag what issues you have encountered and how you have dealt with them. During round 2 you will be required to run your code end/end. This code needs to be provided 12 hours ahead of the Round 2 interview.