State Farm Data Scientist interview questions

Your business partner (e.g., the model consumer) is interested in knowing how the most important feature in the model impacts the model (i.e. the perceived relationship). Identify your favorite binary classification algorithm and explain how you would demonstrate the relationship between the feature and the predicted outcome.

You just received data to start building a classification model. As part of the exploratory analysis, you notice that the dependent variable is very imbalanced. Please name two methods that can help solve the problem of imbalance. Please also provide details for how each method works.

A data scientist was tasked with building a logistic regression model. To help expedite the model build, the data scientist chose to find the top 20 features from a random forest; the features were then added to a main effects only logistic regression model. Surprisingly, the top attribute from the random forest was not statistically significant. Please provide at least 2 reasons.