In the previous blog, we have shared the list of questions which were asked for evaluating communication, confidence, and technical skills (SAS). In the next round, main expectation was to check the candidate for analytics skills.
After the interviewers were comfortable with the technical skills (e.g. SAS in this case), in this round questions were asked to assess the candidate for analytical skills. The questions were mostly based on previous projects - asked to explain any one of your analytics project in detail. Then they asked specific questions about the statistical technique & its applications ( in this case an attrition model). Some of the questions were as follows:
- What was the problem statement? Why were you creating the attrition model? What technique has been used? Initial questions were judge knowledge about requirements understanding.
- Why did you choose logistic regression for this problem? Why can't a linear regression model work in its place?
- At what level did you build the model? (like customer level, branch level etc)
- What was the hypothesis and how did you check whether you can accept the hypothesis?
- Explain the steps used in modeling? (e.g. problem definition, data exploration, data preparation, modeling, validation).
- How did you define model target variable? Explain what steps, like for problem definition how did you arrive at attrition definition? What was the strategy you used etc. How much time did each step take?
- What do you mean by observation period, performance period and lag period? Why is a lag period needed?
- What are the different missing value treatment options available? Which one you used and why? Why is missing value treatment important? What is the treatment for outliers?
- Which variable reduction technique did you use and what all are there?
- What is the equation of a logistic regression model?
- How did you check correlation between your variables? What is multiple regression?
- What are p values and why are they important in determining whether a variable is important for the model or not, explain the concept of p value in non-technical way..
- What are the parameters that determine whether a logistic model is good or not? What is concordance and discordance? How is KS calculated?
- Explain lift chart. What does it mean by the statement ‘The model is giving a lift of 70%’?
Predictive Model Development Training by DnI Institute helps you prepare for all these answer.