Prepare for and practice interview questions from Stellar.Org.

Stellar.Org Interview Questions

Stellar.Org Interview Guides

Machine Learning

Which model would perform better and why?

Booking Regression

Lasso vs Ridge

In which case would you use a bagging algorithm versus a boosting algorithm

Bagging vs Boosting

What metrics would you use to track accuracy and validity of the model

Spam Classifier 

When should you use regularization versus cross-validation

Regularization and Validation

A/B Testing

How would you assess the validity of the result?

Experiment Validity

How would you determine if the new model predicts delivery times better than the old model

Delivery Estimate Model

Button AB Test

A/B Test Power Size

How would you make a control group and test group to account for network effects

Network Experiment Design

Behavioral

Describing a data project and its challenges

Hurdles In Data Projects

Strategically resolving misaligned expectations with stakeholders for a successful project outcome

Stakeholder Communication

How would you prepare and monitor a churn model before and after production deployment?

Model Deployment Pipeline

How would you answer when an Interviewer asks why you applied to their company?

Why Do You Want to Work With Us

Making data-driven insights actionable for those without technical expertise

Simple Explanations

Statistics

Explain what a p-value is to someone who is not technical

P-value to a Layman

Say you are given a dataset of perfectly linearly separable data. What would happen when you run logistic regression?

Perfectly Separable

What are the assumptions of linear regression?

Assumptions of Linear Regression

Explain what an unbiased estimator is and provide a layman-friendly example

Unbiased Estimator

How do you detect and handle correlation between variables in linear regression, and what happens if you ignore it

Correlation in Regression

Product Sense & Metrics

Google Maps Improvement

How would you measure the success of Linkedin’s newsfeed ranking algorithm and approach conflicting success metrics

Ranking Metrics

The role of A/B testing in measuring the success rate of an analytics experiment

Success Measurement

Vision Setting and Execution Strategy

Would you consider adding a payment feature to Facebook Messenger is a good business decision?

Messenger Payments

Let’s say that your company is running a standard control and variant AB test on a feature to increase conversion rates on the landing page. The PM checks the results and finds a .04 p-value.

How would you assess the validity of the result?

Tell me about a data project that didn’t go the way you expected. What did you set out to do, what surprised you, and how did you handle it?

A team wants to A/B test multiple different changes through a sign-up funnel.

For example, on a page, a button is currently red and at the top of the page. They want to see if changing a button from red to blue and/or from the top of the page to the bottom of the page will increase click-through.

How would you set up this test?

Your team wants to run an AB test on the color (red or blue) and position (top/bottom of the page) of a button that links to a promotion.

How many different variants of the test should you run to see the effect that a red button has on the bottom of the page, if currently the button is blue and on the top of the page.

Your team wants to run an AB test on the color (red or blue) and position (top/bottom of the page) of a button that links to a promotion.

How many different variants of the test should you run to see the effect that a red button has on the bottom of the page, if currently the button is blue and on the top of the page.

A/B Test Variants

Let’s say you want to test the close friends feature on Instagram Stories.

How would you make a control group and test group to account for network effects?

What could be a potential risk when Facebook segments the metrics of the user interface change by market or demographic groups?

What could be a potential risk when Facebook segments the metrics of the user interface change by market or demographic groups?

Network Experiment Design I

What is the primary metric that Facebook should monitor if it decides to make the user interface of its posting feature more like Instagram's?

What is the primary metric that Facebook should monitor if it decides to make the user interface of its posting feature more like Instagram's?

Network Experiment Design II

If Facebook changes its user interface to mimic Instagram's, which adverse effect should they anticipate and monitor?

If Facebook changes its user interface to mimic Instagram's, which adverse effect should they anticipate and monitor?

Network Experiment Design III

Let’s say we’re comparing two machine learning algorithms. In which case would you use a bagging algorithm versus a boosting algorithm? 

Give an example of the tradeoffs between the two.

In machine learning, when would you use a bagging algorithm over a boosting algorithm?

In machine learning, when would you use a bagging algorithm over a boosting algorithm?

Bagging vs. Boosting

Let’s say we want to build a new delivery time estimate model for consumers ordering food delivery.

How would you determine if the new model predicts delivery times better than the old model?

When rolling out a new delivery time estimate model in production, what is an important consideration?

When rolling out a new delivery time estimate model in production, what is an important consideration?

Delivery Estimate Model I

What is one way to measure the performance of a new delivery time estimate model for food delivery?

What is one way to measure the performance of a new delivery time estimate model for food delivery?

Delivery Estimate Model II

In an A/B test, how can you check if assignment to the various buckets was truly random?

In an A/B test, which of the following is a common method to determine if the assignment to the test groups was truly random?

In an A/B test, which of the following is a common method to determine if the assignment to the test groups was truly random?

Random Bucketing

How can you check if assignment to the various buckets was truly random

Let’s say we want to build a model to predict booking prices on Airbnb.

Between linear regression and random forest regression, which model would perform better and why?

What are the two key concepts of random forest regression?

What are the two key concepts of random forest regression?

How would you explain what a p-value is to someone who is not technical?

What does a p-value in a statistical test represent?

What does a p-value in a statistical test represent?

P-value to a Layman I

In a statistical test, how does a low p-value (less than 0.05) influence our decision about the null hypothesis?

In a statistical test, how does a low p-value (less than 0.05) influence our decision about the null hypothesis?

P-value to a Layman II

Say you are given a dataset of perfectly linearly separable data.

What would happen when you run logistic regression?

What happens when logistic regression is run on a dataset of perfectly linearly separable data?

What happens when logistic regression is run on a dataset of perfectly linearly separable data?

You’re a data scientist at YouTube. Leadership wants to revisit how the platform measures success, not for a specific new feature, but for the product as a whole given its current state.

How would you define and measure success for YouTube? Identify your primary and secondary metrics, explain your rationale for each, and discuss how you would choose the right thresholds for any engagement-based metrics you propose.

Let’s say that you’re the PM on Google Maps.

<ol>
<li>How would you improve Google Maps?</li>

<li>What metrics would you check to see if your feature improvements are successful?</li>
</ol>

You work as a data engineer at Skyscanner. Your team is responsible for ingesting, transforming, and loading flight and hotel availability data from dozens of external partner APIs into Skyscanner’s real-time search platform. The system must ensure high data freshness, fault tolerance, and scalability as new partners are onboarded.

How would you architect the ETL pipeline to support these requirements? Design a system flow explaining this architecture.

Design a scalable ETL pipeline for ingesting heterogeneous data from Skyscanner's partners.

Skyscanner Partner ETL

Data Pipelines

What’s the difference between Lasso and Ridge Regression?

Under what circumstances is it recommended to use Lasso regression over Ridge regression?

Under what circumstances is it recommended to use Lasso regression over Ridge regression?

Given a <code>users</code> table, write a query to return only its duplicate rows.

Example:

Input:

<code>users</code> table

<table>
<thead>
<tr>
<th>Column</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td><code>id</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>name</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>created_at</code></td>
<td>DATETIME</td>
</tr>
</tbody>
</table>

Duplicate Rows

What are the assumptions of linear regression?

Which assumption of the residuals of the standard linear regression model can not be overcome by increasing the sample size?

Regression assumptions

How would you handle the data preparation for building a machine learning model using imbalanced data?

Addressing imbalanced data in machine learning through carefully prepared techniques.

Data Preparation for Imbalanced Data

Let’s say that you work at a B2B SAAS company that’s interested in testing the pricing of different levels of subscriptions.

Your project manager comes to you and asks you to run a two-week-long A/B test to test an increase in pricing.

How would you approach designing this test? How would you determine whether the increase in pricing is a good business decision?

How would you design a two-week A/B test to determine whether a pricing increase is a good business decision

Testing Price Increase

Let’s say you are tasked with building a spam classifier for emails. Assume that you’ve built a V1 of the model.

What metrics would you use to track accuracy and validity of the model?

Let’s say we’re testing a new UI with the goal to increase conversion rates. We test it by giving the new UI to a random subset of users.

The test variant wins by 5% on the target metric. What would you expect to happen after the new UI is applied to all users? Will the metric actually go up by ~5%, more, or less?

Note: Assume there is no novelty effect.

What would you expect to happen after the new UI is applied to all users?

New UI Effect

Let’s say you have to analyze the results of an AB test.

One variant of the AB test has a sample size of 50K users and the other has a sample size of 200K users.

Given the unbalanced size between the two groups, can you determine if the test will result in bias towards the smaller group?

Given an AB test where one variant has a sample size of 50K users and the other has a sample size of 200K users, what is the likely effect of the unbalanced sizes on the test results?

Given an AB test where one variant has a sample size of 50K users and the other has a sample size of 200K users, what is the likely effect of the unbalanced sizes on the test results?

Sample Size Bias

Can you determine if the test will result in bias towards the smaller group?

What are the key differences between classification models and regression models?

What is the MOST important difference between regression and classification models?

What is the MOST important difference between regression and classification models?

Classification vs Regression

Explain the key differences between classification models and regression models

Classification and Regression

Uber Rider deals with huge amounts of data but at Uber Fleet, the amount of data for experimentation is rather low. Let’s say you run an AB test for Uber Fleet and find out that the distribution is not normal.

What kind of analysis would you run and how would you measure which variant won?

Which of the following is NOT an acceptable way to deal with AB testing on a non-normally distributed population?

Which of the following is NOT an acceptable way to deal with AB testing on a non-normally distributed population?

Non-Normal AB Test

What kind of analysis would you run and how would you measure which variant won

Non-Normal AB Testing

What is an unbiased estimator and can you provide an example for a layman to understand?

What makes an estimator of a population statistic unbiased?

Let’s say you’ve just finished building a churn prediction model for a subscription service, and the team is considering deploying it to automatically trigger retention emails for users at risk of leaving.

How would you ensure this model is ready for production, and what steps would you take to monitor its performance after launch?

When an interviewer asks you a question along the lines of:

<ul>
<li>Why did you apply to our company?</li>
<li>What are you looking for in your next job?</li>
<li>What makes you a good fit for our company?</li>
</ul>

How should you respond?

When asked 'What are you looking for in your next job?' in an interview, how can you tie the company's employee benefits into your response?

When asked 'What are you looking for in your next job?' in an interview, how can you tie the company's employee benefits into your response?

Why Do You Want to Work With Us I

How can company values be used effectively in an interview when asked 'What makes you a good fit for our company?'

How can company values be used effectively in an interview when asked 'What makes you a good fit for our company?'

Why Do You Want to Work With Us II

When responding to the question 'Why did you apply to our company?' during an interview, what aspect should you highlight?

When responding to the question 'Why did you apply to our company?' during an interview, what aspect should you highlight?

Why Do You Want to Work With Us III

What are time series models? Why do we need them when we have less complicated regression models?

What can happen if a traditional regression model is used on data that has autocorrelation?

What can happen if a traditional regression model is used on data that has autocorrelation?

Why Do We Need Time Series Models? I

Why can't standard regression models be used effectively for data points where current values are influenced by past values?

Why can't standard regression models be used effectively for data points where current values are influenced by past values?

Why Do We Need Time Series Models? II

Explain what time series models are and why they’re needed instead of simpler regression models

Why Do We Need Time Series Models?

Forecasting & Time Series

Build a logistic regression model from scratch, with the following conditions:

<ul>
<li>Return the parameters of the regression</li>
<li>Do not include an intercept term</li>
<li>Use basic gradient descent (with newton’s method) as your optimization method and the log-likelihood as your loss function.</li>
<li>Don’t include a penalty term.</li>
<li>You may use numpy and pandas but NOT scikit-learn</li>
</ul>

Example:

Input:

<pre tabindex="0" class="chroma"><code>step_size = 0.02
max_steps = 100
starting_point = np.asarray(data.drop(&#39;Target&#39;).mean())
print(data)
...
 Var1 Var2 Var3 Target
0 0.439326 -0.662463 0.027418 1
1 -1.363880 -1.303938 -1.595875 1
2 -1.710828 -2.473115 -0.968141 1
3 -1.620589 1.321073 1.608265 0
4 -0.315593 -1.655797 0.126860 0
.. ... ... ... ...
95 -0.663844 0.225382 0.650974 0
96 1.479807 0.322281 0.501116 0
97 0.510666 -0.745397 -0.239914 0
98 -0.769290 1.348070 -0.059427 0
99 2.018888 -0.945268 -0.851626 0

[100 rows x 4 columns]
</code></pre>

Output:

<pre tabindex="0" class="chroma"><code>def logistic_regression(data,starting_point,step_size,steps) -&gt; [0.35155765,-0.30863254,-0.13125343]
</code></pre>

Implement logistic regression from scratch in code

Logistic Regression from Scratch

Write a function <code>fib</code> which takes an integer <code>n</code> and returns the nth fibonacci number. For this question, <code>fib</code> MUST be recursively defined.

Note: <code>for</code> and <code>while</code> keywords are disabled for this question, as well as <code>functools</code>. Any workarounds around this restriction is prohibited.

Example:

Input:

<pre tabindex="0" class="chroma"><code>n = 4
</code></pre>

Output:

<pre tabindex="0" class="chroma"><code>def fib(n) -&gt; 3
# the sequence is [1, 1, 2, 3, 5, 8...]
</code></pre>

Implement a dynamic recursive fibonacci function.

Impossibly Iterative Fibonacci

Data Structures & Algorithms

Imagine you work at a major credit card company and are given a dataset of 600,000 credit card transactions to build a fraud detection model.

How would you approach this task? Write your answer in the comments.

Calculate Moving Average	SQL	Easy
Predict Customer Churn	Machine Learning	Medium
A/B Test Significance	Statistics	Medium
Optimize Query Performance	SQL	Hard
Feature Importance Analysis	Machine Learning	Medium
Clean Missing Data	Python	Easy
Neural Network Architecture	Deep Learning	Hard
Calculate Cohort Retention	SQL	Medium
Bayesian Probability	Statistics	Easy
Recommend Similar Products	Machine Learning	Hard

Stellar.Org Interview Questions

Stellar.Org Interview Guides

Stellar.Org Interview Questions

Challenge

Discussion & Interview Experiences

Discussion & Interview Experiences