Application

I interviewed at Millenium London in an onsite round for the role of Quant Researcher.

Interview

Duration: 70 Mins

The interview was long and detailed, with questions ranging from easy to hard. The interviewer was friendly and gave me enough time to think and solve questions correctly. The interview started with some basic questions from the resume, followed by questions on statistics, and ended with deep discussions on the resume.

Interview Questions

Why are you interested in trading?
Explain your trading strategy (based on resume). What were the KPIs for evaluating your strategy?
Explain steps for data cleaning of time series.
How to backtest a strategy?
You are given 100 features, and you have to select K features from them; how will you proceed?
What is the difference between bagging and boosting?
The correlation of X and Y is 0. What happens to the slope of the linear regression line fit between X and Y?
Given an example of the Distribution of X and Y where Corr(X,Y) = 0, but X and Y depend on each other.
Derivation of Covariance (with and without independence)
How to reduce overfitting in a random forest?
[Hardest Question] Random Forest may give different feature importance when you change the random seed. How to evaluate which random seed is better?
Example: Let random_seed = 1234, the model may output top features as (A,B,C,D), but if we change random_seed to 1001, the top features are (B,E,D,A). Which random_seed to select?