Amazon Data Science Interview –

Here we gonna talk about the interview process in Amazon Data Science interview, Amazon is a  product based company so have multiple rounds of interviews, let’s talk one by one of each and  every round and then we will discuss some of the most often questions asked in Amazon Interview  for Data Science Roles.  

 

 

There are majorly 4 rounds to hire a Data Scientist/Machine Learning Engineer.  Round 1- Screening Round(Online-Test)  

This round could be a simple programming and aptitude to test whether a candidate is comfortable  in programming, programming language is not a always a choice as in many other product based  companies.  

For Free, Demo classes Call: 8983120543
Registration Link: Click Here!

What Topics you should prepare to qualify this round?  

Data Structure And Algorithms –  

Programming questions from Arrays, LinkedList, Binary Search Tree, Graphs, Hash tables are very  common and often programming questions revolve around these topics.  

To prepare data structures well just go to leetcode or hackerrank and just do around 250 to 300  questions and understand their basic mechanism, this will take around 3 to 6 months, depending  upon the time you are giving to coding practise. 

Aptitude  –

There has been a lot of online material on building your aptitude, so take any book or refer any  website, practise it for a month and you are prepared to clear any aptitude interview.  

Round-2 (Face-Time)  –

If you cleared the screening round, it’s time to have a face to face interview. This interview will  mostly be on your portfolio, your resume, your projects and all the technical stuff you have used in  your projects.  

You should be very well versed with your resume and portfolio, practise in advance about all the  questions interview can ask and go into the depths of topic. 

In any product based company the interviewer will check your breadth and depth of topics so it  should be prepared accordingly.  

Make sure be very through in mathematics of all the algorithms, the interviewer will go into the  depths and breadths of the algorithms you have used in projects. 

For example, if you have used Logistic Regression, then make sure you know in and out of logistics  regression.  

For Free, Demo classes Call: 8983120543
Registration Link: Click Here!

Some of the questions could be:  

1. What is Logistic Regression?  
2. Is Logistic Regression a Regression or Classification Technique?  
3. Can we do multi-class classification using logistic regression?  
4. What is Sigmoid Function?  
5. What is Squashing and why it is needed?  
6. How Logistic Regression is handling imbalanced data?  
7. What is Regularisation?  
8. Why do we need Regularisation?  
9. What is the difference between L1 and L2 Regulariser?  

This round also can go in different directions depending upon your answers like, if you talk  about probability and statistics , they will ask you more about questions like :-  

1. What is Hypothesis Test?  
2. What is Confidence Interval?  
3. What is P value?  
4. What is Q-Q Test?  
5. What is K-S Test? 
6. What is Permutation-Resampling test?  
7. What is Gaussian Distribution and Power Law Distribution?  
8. What is A/B testing?  
9. What are the types of Sampling?  
10. What is Central Limit Theorem?  

Round-3 (Problem-Statement Round)  

Now if you have cleared second round, this round will be mostly on the problem statement amazon  may working on.  

Lets say Amazon is working on Recommendation systems, then they can give you a problem  statement very similar to recommendation system and they can ask you questions like-:  

1. How to build a Recommendation System?  
2. What is content based Recommendation System?  
3. What is Matrix Factorisation?  
4. What is SVD(Singular Value Decomposition)?  
5. What is PCA(Principle Component Analysis)?  
6. What is collaborative filtering?  
7. How Netflix is Recommending Movies/Series to you?  
8. How Spotify is Recommending songs to you?  
9. What is a Cold Start Problem?  
10. What are the algorithms widely used in industry to solve the Recommendation System  Problem?  

Different problem statements could be given to you according to the problem company is solving.  

Round-4 (HR-Round)  

The last round will be your HR round which will be focused on your personality, your salary,  company’s environment.  

This will be a typical HR round so i am not adding questions in order to keep this blog on technical  side. 

For Free, Demo classes Call: 8983120543
Registration Link: Click Here!

SOME OTHER QUESTIONS WHICH WERE ASKED IN PRODUCT BASED  COMPANIES…  

1. If your model is giving you 80% accuracy on test data, will you deploy this model, if yes  then why? if no then why?  
2. What you mean by Overfitting of Model?  
3. What is Curse of dimensionality?  
4. What are the ways to handle outliers in data?  
5. What do you mean by Imbalanced data, how to solve this problem?  
6. If a couple has 2 children, one is a boy. What is the probability that other kid is also a boy?  7. What is Local Outlier Factor?  
7. What is the assumption of KNN?  
8. Can a Random Forest Model be overfit, if yes, give the examples.  
9. What is Ensembles ?  

I hope this blog will help you to prepare for product based companies, keep hustling for your dream  job, keep applying , keep trying and keep failing. One day you will crack your favorite company’s  interview.  

Thanks for reading.

Author: