Prepare for and practice interview questions from Belcan Engineering.

Belcan Engineering Interview Questions

Belcan Engineering Interview Guides

Machine Learning

Lasso vs Ridge

Why would one algorithm generate different success rates with the same dataset?

Same Algorithm Different Success

How would you build a model to detect if a post on a marketplace is talking about selling a gun?

Detecting Firearm Sales

Classification and Regression

Xgboost vs Random Forest

Behavioral

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

Why Do You Want to Work With Us

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 to present complex data insights with clarity and adaptability tailored to a specific audience

Presentations and Insights

What do you tell an interviewer when they ask you what your strengths and weaknesses are?

Your Strengths and Weaknesses

Calculate the first touch attribution for each `user_id` that converted.

First Touch Attribution

Project Budget Error

Employee Project Budgets

Count the number of users that like each user

Liker's Likers

How would you design a data warehouse for a e-commerce company looking to expand internationally?

International e-Commerce Warehouse

Data Structures & Algorithms

String Mapping

Swap Variables

This question requires the implementation of the Fibonacci sequence using three different methods: recursively, iteratively, and using memoization.

Implementing the Fibonacci Sequence in Three Different Methods

Write a class 'SinglyLinkedList' that implements the interface required for a Singly Linked List.

Singly Linked List

Query Optimization

How would you diagnose and speed up a slow SQL query when system metrics look healthy?

Slow SQL Query

When an interviewer asks a question along the lines of:

<ul>
<li>What would your current manager say about you? What constructive criticisms might he give?</li>
<li>What are your three biggest strengths and weaknesses you have identified in yourself?</li>
</ul>

How would you respond?

When asked about your strengths in an interview, what is an effective way to respond?

When asked about your strengths in an interview, what is an effective way to respond?

Your Strengths and Weaknesses I

Which of the following is an acceptable strategy when discussing weaknesses in an interview?

Which of the following is an acceptable strategy when discussing weaknesses in an interview?

Your Strengths and Weaknesses II

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

Describe a data project you worked on. What were some of the challenges you faced?

Talk about a time when you had trouble communicating with stakeholders. How were you able to overcome it?

Imagine you are asked to build a machine learning model to decide new loan approvals for a
financial firm. You ask the data department in the company for a subset of data to get started
working on the problem. The data includes different features about applicants such as age,
occupation, zip code, height, number of children, favorite color, etc. You decide to build
multiple machine learning models to test out different ideas before settling on the best one.

How would you explain the bias-variance tradeoff with regards to
building and choosing a model to use?

Bias vs. Variance Tradeoff

Explain the difference between the XGBoost and random forest algorithms and give an example where you would use one over the other.

Why might a data scientist choose to use XGBoost instead of Random Forest for a particular machine learning task?

Why might a data scientist choose to use XGBoost instead of Random Forest for a particular machine learning task?

XGBoost vs Random Forest: Choice

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

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?

What’s the relationship between PCA and K-means clustering?

What does the variable “k” in k-means clustering refer to?

What does the variable "k" in k-means clustering refer to?

Input of K-means

PCA and K-Means

The Fibonacci sequence is a series of numbers in which each number is the sum of the two preceding ones, usually starting with 0 and 1. It is often used in algorithm examples, and is defined by the following formula: F(n) = F(n-1) + F(n-2), with F(0) = 0 and F(1) = 1.

Your task is to implement the Fibonacci algorithm in three different methods:
1. Recursively
2. Iteratively
3. Using Memoization

Example 1:

Input:

<pre tabindex="0" class="chroma"><code>n = 5
</code></pre>

Output:

<pre tabindex="0" class="chroma"><code>fibonacci(n) -&gt; 5
</code></pre>

Example 2:

Input:

<pre tabindex="0" class="chroma"><code>n = 10
</code></pre>

Output:

<pre tabindex="0" class="chroma"><code>fibonacci(n) -&gt; 55
</code></pre>

The Fibonacci sequence starts as follows: 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55…

Let’s say you’re a data engineer at Fidelity Investments, and you’re running a SQL query on a cloud-based data warehouse. All cluster resources and network health metrics look normal, but the query is still taking over 10 minutes to complete.

How would you go about diagnosing and improving the performance of this query?

What does the backpropagation algorithm do in the context of neural networks? What is the informal intuition behind the algorithm? What are some drawbacks of the algorithm compared to other optimization methods?

Bonus: Formally derive the backpropagation algorithm and prove that it does what it claims to do.

Backpropagation Explanation

You are given a dictionary with two keys <code>a</code> and <code>b</code> that hold integers as their values.

Without declaring any other variable, swap the value of <code>a</code> with the value of <code>b</code> and vice versa.

Note: Return the dictionary after editing it.

Example:

Input:

<pre tabindex="0" class="chroma"><code>numbers = {
 &#39;a&#39;:3,
 &#39;b&#39;:4
}
</code></pre>

Output:

<pre tabindex="0" class="chroma"><code>def swap_values(numbers) -&gt; {&#39;a&#39;:4,&#39;b&#39;:3}
</code></pre>

Let’s say you work for an e-commerce company. Vendors can send products to the company’s warehouse to be listed on the website. Users are able to order any in-stock products and submit returns for refunds if they’re not satisfied.

The front end of the website includes a vendor portal that provides sales data in daily, weekly, monthly, quarterly, and yearly intervals.

The company wants to expand worldwide. They put you in charge of designing its end-to-end architecture, so you need to know what significant factors you’ll need to consider. What clarifying questions would you ask?

What kind of end-to-end architecture would you design for this company (both for ETL and reporting)?

Let’s say that you work on the revenue forecasting team at a company like Facebook.

An executive comes to you asking about how much revenue Facebook will make in the coming year.

How would you forecast revenue for the next year?

Forecasting New Year Revenue

Forecasting & Time Series

How comfortable are you presenting your insights?

As a data scientist at a real estate analytics startup, you’re tasked with building a predictive model for home prices in a large metropolitan area. During your exploratory data analysis, you notice that the <code>home_prices</code> column in your dataset is heavily right-skewed (i.e., there are many lower-priced homes and a few extremely high-priced ones).

How might this skewness impact your modeling approach, and what steps would you consider to address it before training your model?

Bonus: If, instead, you discover that the target variable (<code>home_prices</code>) is heavily left-skewed (i.e., there are many high-priced homes and a few low-priced outliers), what would you do differently, if anything?

You're building a model to predict home prices and you see that the prices in your training data are right-skewed. How would you handle this situation?

Skewed Pricing

Singly-linked lists are a type of linked list where nodes have only one pointer: A pointer to the next element. Create a class <code>SinglyLinkedList</code> along with its methods:

<ul>
<li><code>add_head(element)</code> - adds an element to the head of the list. This becomes the new head of the linked list.</li>
<li><code>add_tail(element)</code> - adds an element to the tail of the list. This becomes the new tail of the linked list.</li>
<li><code>remove_head() -&gt; element</code> - removes the head and returns the value of the removed element.</li>
<li><code>remove_tail() -&gt; element</code> - removes the tail and returns the value of the removed element.</li>
<li><code>__contains__(item) -&gt; bool</code> - checks if the item is inside the list. Returns <code>True</code> if it is inside the list and returns <code>False</code> if not. This method is implemented to allow the use of Python’s <code>in</code> operator.</li>
<li><code>__getitem__(index)</code> - returns the element at the specified index. Raises an <code>IndexError</code> if the index is out of bounds. This method is implemented to allow the use of Python’s <code>[]</code> operator.</li>
<li><code>__len__()</code> - returns the length of the <code>SinglyLinkedList</code>.</li>
</ul>

Why would the same machine learning algorithm generate different success rates using the same dataset?

Note: When they ask us an ambiguous question, we need to gather context and restate it in a way that’s clear for us to answer.

Given two strings, <code>string1</code> and <code>string2</code>, write a function <code>str_map</code> to determine if there exists a one-to-one correspondence (bijection) between the characters of <code>string1</code> and <code>string2</code>.

For the two strings, our correspondence must be between characters in the same position/index.

Example 1:

Input:

<pre tabindex="0" class="chroma"><code>string1 = &#39;qwe&#39;
string2 = &#39;asd&#39;

string_map(string1, string2) == True

# q = a, w = s, and e = d
</code></pre>

Example 2:

Input:

<pre tabindex="0" class="chroma"><code>string1 = &#39;donut&#39;
string2 = &#39;fatty&#39;

string_map(string1, string2) == False
# cannot map two distinct characters to two equal characters
</code></pre>

Example 3:

Input:

<pre tabindex="0" class="chroma"><code>string1 = &#39;enemy&#39;
string2 = &#39;enemy&#39;

string_map(string1, string2) == True
# there exists a one-to-one correspondence between equivalent strings
</code></pre>

Example 4:

Input:

<pre tabindex="0" class="chroma"><code>string1 = &#39;enemy&#39;
string2 = &#39;ymene&#39;

string_map(string1, string2) == False
# since our correspondence must be between characters of the same index, this case returns &#39;False&#39; as we must map e = y AND e = e
</code></pre>

We’re given two tables. One is named <code>projects</code> and the other maps employees to the projects they’re working on.

Write a query to get the top five most expensive projects by budget to employee count ratio.

Note: Exclude projects with 0 employees. Assume each employee works on only one project.

<code>projects</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>title</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>state_date</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>end_date</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>budget</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
<code>employee_projects</code> table

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

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

<tr>
<td><code>employee_id</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
Output:

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

<tbody>
<tr>
<td><code>title</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>budget_per_employee</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>

Let’s say you are designing a marketplace for your website.

Selling firearms is prohibited by your website’s Terms of Service Agreement (not to mention the laws of your country). To this end, you want to create a system that can automatically detect if a listing on the marketplace is selling a gun.

How would you go about doing this?

While designing a system to detect firearms listings on a marketplace website, what metrics could be used to minimize both false positives and false negatives?

While designing a system to detect firearms listings on a marketplace website, what metrics could be used to minimize both false positives and false negatives?

There are two tables. One table is called <code>swipes</code> that holds a row for every Tinder swipe and contains a boolean column that determines if the swipe was a right or left swipe called <code>is_right_swipe</code>. The second is a table named <code>variants</code> that determines which user has which variant of an AB test.

Write a SQL query to output the average number of right swipes for two different variants of a feed ranking algorithm by comparing users that have swiped 10, 50, and 100 swipes in a <code>feed_change</code> experiment.

Note: Users have to have swiped at least 10 times to be included in the subset of users to analyze the mean number of right swipes.

Example:

Input:

<code>variants</code> table

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

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

<tr>
<td><code>experiment</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>variant</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>user_id</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
<code>swipes</code> table

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

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

<tr>
<td><code>user_id</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>swiped_user_id</code></td>
<td>INTEGER</td>
</tr>

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

<tr>
<td><code>is_right_swipe</code></td>
<td>BOOLEAN</td>
</tr>
</tbody>
</table>
Output:

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

<tbody>
<tr>
<td><code>variant</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>mean_right_swipes</code></td>
<td>FLOAT</td>
</tr>

<tr>
<td><code>swipe_threshold</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>num_users</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>

Write a SQL query to find the average number of right swipes for different ranking algorithms.

Swipe Precision

A/B Testing

The schema below is for a retail online shopping company consisting of two tables, <code>attribution</code> and <code>user_sessions</code>.

<ul>
<li>The attribution table logs a session visit for each row.</li>

<li>If <code>conversion</code> is <code>true</code>, then the user converted to buying on that session.</li>

<li>The <code>channel</code> column represents which advertising platform the user was attributed to for that specific session.</li>

<li>Lastly the <code>user_sessions</code> table maps many to one session visits back to one user.</li>
</ul>

First touch attribution is defined as the channel with which the converted user was associated when they first discovered the website.

Calculate the first touch attribution for each <code>user_id</code> that converted. 

Example:

Input:

<code>attribution</code> table

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

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

<tr>
<td><code>channel</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>conversion</code></td>
<td>BOOLEAN</td>
</tr>
</tbody>
</table>
<code>user_sessions</code> table

<table>
<thead>
<tr>
<th>column</th>
<th>type</th>
</tr>
</thead>

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

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

<tr>
<td><code>user_id</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
Example output:

<table>
<thead>
<tr>
<th><code>user_id</code></th>
<th><code>channel</code></th>
</tr>
</thead>

<tbody>
<tr>
<td>123</td>
<td>facebook</td>
</tr>

<tr>
<td>145</td>
<td>google</td>
</tr>

<tr>
<td>153</td>
<td>facebook</td>
</tr>

<tr>
<td>172</td>
<td>organic</td>
</tr>

<tr>
<td>173</td>
<td>email</td>
</tr>
</tbody>
</table>

We’re given two tables. One is named <code>projects</code> and the other maps employees to the projects they’re working on. 

We want to select the five most expensive projects by budget to employee count ratio. But let’s say that we’ve found a bug where there exist duplicate rows in the <code>employee_projects</code> table.

Write a query to account for the error and select the top five most expensive projects by budget to employee count ratio.

<h3>Schema</h3>

Input:

<code>projects</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>title</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>state_date</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>end_date</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>budget</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
<code>employee_projects</code> table

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

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

<tr>
<td><code>employee_id</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
Output:

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

<tbody>
<tr>
<td><code>title</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>budget_per_employee</code></td>
<td>FLOAT</td>
</tr>
</tbody>
</table>

<h3>Example</h3>

Input:

<code>projects</code> table

<table>
<thead>
<tr>
<th>id</th>
<th>title</th>
<th>start_date</th>
<th>end_date</th>
<th>budget</th>
</tr>
</thead>

<tbody>
<tr>
<td>1</td>
<td>party</td>
<td>2006-04-17 00:00:00</td>
<td>2006-05-19 00:00:00</td>
<td>25875</td>
</tr>

<tr>
<td>2</td>
<td>diversity</td>
<td>2017-08-09 00:00:00</td>
<td>2017-09-18 00:00:00</td>
<td>77867</td>
</tr>

<tr>
<td>3</td>
<td>integration</td>
<td>2005-06-29 00:00:00</td>
<td>2005-07-19 00:00:00</td>
<td>75987</td>
</tr>

<tr>
<td>4</td>
<td>testing</td>
<td>2009-01-12 00:00:00</td>
<td>2009-05-03 00:00:00</td>
<td>35946</td>
</tr>

<tr>
<td>5</td>
<td>launch</td>
<td>2005-05-02 00:00:00</td>
<td>2005-10-16 00:00:00</td>
<td>66292</td>
</tr>

<tr>
<td>6</td>
<td>meet</td>
<td>2001-09-09 00:00:00</td>
<td>2002-04-04 00:00:00</td>
<td>71243</td>
</tr>

<tr>
<td>7</td>
<td>payroll</td>
<td>2019-09-25 00:00:00</td>
<td>2020-05-05 00:00:00</td>
<td>97071</td>
</tr>

<tr>
<td>8</td>
<td>admin</td>
<td>2000-07-08 00:00:00</td>
<td>2000-09-10 00:00:00</td>
<td>24000</td>
</tr>

<tr>
<td>9</td>
<td>petronas</td>
<td>2016-07-01 00:00:00</td>
<td>2016-11-28 00:00:00</td>
<td>37088</td>
</tr>

<tr>
<td>10</td>
<td>bel</td>
<td>2005-08-03 00:00:00</td>
<td>2006-04-29 00:00:00</td>
<td>61937</td>
</tr>
</tbody>
</table>
<code>employee_projects</code> table

<table>
<thead>
<tr>
<th>project_id</th>
<th>employee_id</th>
</tr>
</thead>

<tbody>
<tr>
<td>1</td>
<td>7</td>
</tr>

<tr>
<td>1</td>
<td>9</td>
</tr>

<tr>
<td>2</td>
<td>6</td>
</tr>

<tr>
<td>3</td>
<td>1</td>
</tr>

<tr>
<td>3</td>
<td>2</td>
</tr>

<tr>
<td>4</td>
<td>5</td>
</tr>

<tr>
<td>5</td>
<td>3</td>
</tr>

<tr>
<td>5</td>
<td>4</td>
</tr>

<tr>
<td>6</td>
<td>8</td>
</tr>

<tr>
<td>8</td>
<td>10</td>
</tr>

<tr>
<td>9</td>
<td>11</td>
</tr>

<tr>
<td>10</td>
<td>12</td>
</tr>

<tr>
<td>5</td>
<td>5</td>
</tr>

<tr>
<td>9</td>
<td>11</td>
</tr>

<tr>
<td>6</td>
<td>8</td>
</tr>

<tr>
<td>4</td>
<td>5</td>
</tr>
</tbody>
</table>
Output:

<table>
<thead>
<tr>
<th>title</th>
<th>budget_per_employee</th>
</tr>
</thead>

<tbody>
<tr>
<td>diversity</td>
<td>77867</td>
</tr>

<tr>
<td>meet</td>
<td>71243</td>
</tr>

<tr>
<td>bel</td>
<td>61937</td>
</tr>

<tr>
<td>integration</td>
<td>37993.5</td>
</tr>

<tr>
<td>petronas</td>
<td>37088</td>
</tr>
</tbody>
</table>

A dating website’s schema is represented by a table of people that like other people. The table has three columns. One column is the <code>user_id</code>, another column is the <code>liker_id</code> which is the <code>user_id</code> of the user doing the liking, and the last column is the <code>date time</code> that the like occurred.

Write a query to count the number of liker’s likers (the users that like the likers) if the liker has one.

Example:

Input:

<code>likes</code> table

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

<tbody>
<tr>
<td><code>user_id</code></td>
<td>VARCHAR</td>
</tr>

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

<tr>
<td><code>liker_id</code></td>
<td>VARCHAR</td>
</tr>
</tbody>
</table>
Output:

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

<tbody>
<tr>
<td><code>user</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>count</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>

Let’s say you’re a senior data analyst and your team is tasked with addressing data quality issues for an airline company. Customer records, booking details, and flight schedules often have missing fields, duplicate entries, or incorrect data.

How would you approach improving the quality of this data and ensuring it remains clean and reliable over time?

How would you approach improving the quality of airline data?

Addressing Data Quality Issues

Analytics

Let’s say we manage a factory where the current manual production line employs 50 workers, and we have the option to upgrade to a robotic system that can produce parts at 3 times the current rate.

How would you approach deciding between upgrading to robotics or keeping the manual line, if our goals are to minimize total production time while also maintaining high employee satisfaction?

How would you balance production speed and employee satisfaction when considering a switch to robotics?

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

Belcan Engineering Interview Questions

Belcan Engineering Interview Guides

Belcan Engineering Interview Questions

Challenge

Belcan Engineering Salaries by Position

Discussion & Interview Experiences

Discussion & Interview Experiences