Interview Practice Questions

Implement a data structure with insert, remove, and getRandom — all in O(1). getRandom returns a uniformly random element.

Given server logs [timestamp, server_id, start/end], find the time interval with maximum concurrent active servers. Handle 10⁸ logs efficiently.

What is the time complexity of this algorithm?

Could you write something into the object?

To clarify why I am asking: have you solved LeetCode problems, at least easy ones?

What is your algorithms experience? Have you practiced LeetCode, and do you write in Python?

What approaches are available for encoding categorical features?

Are there methods that do not depend on already having a trained model?

Where does the information encoding take place?

Instead of doing everything in a function, computing it, and putting it all into an array?

Or do we specifically need to output an array?

Under what conditions will line 17, after the loop, execute? Will it execute when we reach the last element?

Is the algorithm O(n) because we iterate over every element in the list?

What do we need the machine learning algorithm to determine as its output?

Logistic regression, because it is a relatively simple algorithm.

The optimizer determines how we improve our model based on this loss function.

I cannot remember the NumPy method or function that computes Euclidean distance.

Describe your professional experience in ML and computer vision.

Which ML domains do you work in?

Describe your current role and main responsibilities.