Top Machine Learning Interview Questions

Formulate AI
2 min readMay 1, 2021

Many a times when we start preparing for our machine learning interviews, we wanted a list of questions to check out our confidence level and identify areas to brush up.

By seeing this demand, formulators have collected a set of top ’N’ questions which is asked by many of the companies and these are pretty common. Idea of this blog is to just make sure that you skim through all these questions and if you don’t know any thing definitely brush it up

Top ’N’ Questions in ML

1.How do you structure machine learning projects?

2.What is the difference between the Statistical model and the machine learning model?

3.If you have outliers and missing values in the data, which one to handle first and why?

4.Tell me assumptions of Linear Regression?

5.Why Logistic regression is called Regression?

6.How Time series model is different from the regression model?

7.Can you explain these terms Confusion matrix, precision, recall, f1-score, specificity, and sensitivity of the model?

8.How to handle imbalanced data?

9.What is Type1 & type 2 error, which one is important and why?

10.How will you create a clustering algorithm (like K-Means) from scratch?

11.Different ways of achieving scaling?

12.What is the Silhouette Score and how do you calculate it?

13.What is the AUC curve?

14.What is alpha in exponential smoothing?

15.How many regularisation techniques you have implemented?

16.What is PCA?

17.Difference between PCA & LDA?

18.What are the hyperparameters used in SVM, decision tree?

A score 14 out of 18 is considered that you are ready to take you next interview.

In case if you feel not sure about any concepts and would like to have 1:1 mentoring or a group of data scientists helping you, we would love to help you with our True Mentor Platform.

You can register free at True Mentor Platform and explore the groups and ask your doubts using hashtags like #help. Your queries will be mostly solved in the common group itself and in few case it can require 1:1 mentoring with the supporting mentor (private room chats). We have 700+ mentors already onboarded hence looking forward to see you at True Mentor.

All the best for your next interview. Keep Learning & Keep Formulating!

--

--