Data Scientist Interview

Question 1

What is variance?

Answer 1

Variance measures data spread by averaging squared differences from the mean. Higher variance means more spread.

Variance is a key concept in statistics that helps in understanding the dispersion of data points in a dataset.

Question 2

Difference between mean and median?

Answer 2

Mean is the average. Median is the middle value. Median is less affected by outliers.

The mean can be skewed by extreme values, while the median provides a better measure of central tendency in such cases.

Question 3

How do you detect outliers?

Answer 3

Use Z-score (>3), IQR (1.5*IQR rule), or visualize with boxplots and histograms.

Outlier detection is crucial for data cleaning and ensuring the accuracy of statistical analyses.

Question 4

Explain Bayes’ Theorem

Answer 4

It updates probabilities based on new evidence: P(A|B) = P(B|A) * P(A) / P(B).

Bayes’ Theorem is fundamental in statistics and machine learning for making inferences based on prior knowledge.

Question 5

When to use T-test vs. Chi-square vs. ANOVA?

Answer 5

T-test: 2 means, Chi-square: categorical independence, ANOVA: 3+ means.

These tests are used to compare means or distributions in various scenarios, depending on the number of groups and data types.

Question 6

What’s a p-value?

Answer 6

The probability of getting the observed result if the null hypothesis is true. p < 0.05 is significant.

P-values are used in hypothesis testing to determine the strength of the evidence against the null hypothesis.

Question 7

What is Multicollinearity, and how to detect it?

Answer 7

When features are highly correlated. Detect using VIF (>5) or correlation matrices.

Multicollinearity can affect the stability of regression coefficients and complicate interpretations.

Question 8

How do you handle imbalanced datasets?

Answer 8

Resampling, class weighting, better metrics like F1-score, and tree-based models.

Addressing imbalance is essential for improving model performance and ensuring fair predictions.

Question 9

What is a window function?

Answer 9

A function that performs calculations across a specific row window, like RANK or LEAD/LAG.

Question 10

Difference between RANK, DENSE_RANK, and ROW_NUMBER?

Answer 10

RANK skips ranks on ties, DENSE_RANK does not, ROW_NUMBER gives unique numbers.

Question 11

What are CTEs and subqueries?

Answer 11

CTEs (Common Table Expression) improve readability and can be recursive. Subqueries are nested queries inside another query.

Question 12

Difference between INNER JOIN and LEFT JOIN?

Answer 12

INNER JOIN keeps only matching rows, LEFT JOIN keeps all left table rows.

Question 13

What is vectorization in Python?

Answer 13

Using NumPy or pandas operations instead of loops for faster execution.

Question 14

What is the difference between a list and a tuple?

Answer 14

Lists are mutable, tuples are immutable and faster.

Question 15

What is a hash table?

Answer 15

A data structure that maps keys to values using a hash function for fast lookup.

Question 16

Difference between BFS and DFS?

Answer 16

BFS (Breadth First Search) explores level by level, DFS (Depth First Search) goes deep first.

Question 17

What is overfitting?

Answer 17

A model that learns noise instead of patterns, performing poorly on new data.

Question 18

Difference between supervised and unsupervised learning?

Answer 18

Supervised uses labeled data, unsupervised finds patterns in unlabeled data.

Question 19

What is the difference between bagging and boosting?

Answer 19

Bagging reduces variance (Random Forest), boosting corrects errors (XGBoost).

Question 20

Difference between logistic regression and SVM?

Answer 20

Logistic regression is simpler, SVM works better for complex decision boundaries.

Question 21

Why use ReLU in neural networks?

Answer 21

It avoids vanishing gradients and speeds up training.

Question 22

What is backpropagation?

Answer 22

A method to update weights using gradients of the loss function.

Question 23

Difference between CNNs and RNNs?

Answer 23

CNNs handle images, RNNs handle sequential data like text or time series.

Question 24

What is dropout in neural networks?

Answer 24

Randomly deactivates neurons during training to prevent overfitting.

Question 25

Difference between batch normalization and layer normalization?

Answer 25

Batch norm normalizes per feature (better for CNNs and dense networks) Layer norm normalizes per sample (RNNs, transformers, and small batch sizes) Both stabilize and speed up training in deep neural networks by normalizing activations.

Question 26

What is MapReduce?

Answer 26

A framework for processing big data by splitting tasks into map and reduce steps. Map Phase: Convert words into key-value pairs [('apple', 1), ('banana', 1), ('apple', 1), ('banana', 1), ('orange', 1), ('apple', 1)] Shuffle & Sort: Group by word {'apple': [1, 1, 1], 'banana': [1, 1], 'orange': [1]} Reduce Phase: Sum up values for each word. {'apple': 3, 'banana': 2, 'orange': 1}

Question 27

What is a data pipeline?

Answer 27

A sequence of data processing steps, often automated.

Question 28

Difference between ETL and ELT?

Answer 28

ETL transforms data before loading, ELT loads first, then transforms.

Question 29

What is a NoSQL database?

Answer 29

A non-relational database like MongoDB that stores data in flexible formats.

Question 30

What metrics would you track for a new product?

Answer 30

Retention, churn, conversion rate, and user engagement.

Question 31

How do you improve a recommendation system?

Answer 31

Use collaborative filtering, content-based filtering, or hybrid approaches.

Question 32

Difference between DAU, MAU, and retention rate?

Answer 32

DAU: daily active users, MAU: monthly active users, retention: returning users.

Question 33

How do you measure A/B test success?

Answer 33

Low p-value Significant lift in key metric (ie conversion rate, revenue, engagement, etc). - Lift is improvement in test group (new version w changes being tested, B) compared to control group (og version, A)

Question 34

How do you find the second-highest salary in SQL?

Answer 34

`SELECT MAX(salary) FROM employees WHERE salary < (SELECT MAX(salary) FROM employees);`

Question 35

How do you count the number of customers per country in SQL?

Answer 35

`SELECT country, COUNT(*) FROM customers GROUP BY country;`

Question 36

How do you find customers who placed more than 2 orders in SQL?

Answer 36

`SELECT customer_id FROM orders GROUP BY customer_id HAVING COUNT(order_id) > 2;`

Question 37

How do you calculate the rolling average of sales in the last 3 months in SQL?

Answer 37

`SELECT month, sales, AVG(sales) OVER (ORDER BY month ROWS BETWEEN 2 PRECEDING AND CURRENT ROW) FROM sales_data;`

Question 38

How do you find the top 3 highest salaries per department in SQL?

Answer 38

`SELECT employee_name, department, salary, RANK() OVER (PARTITION BY department ORDER BY salary DESC) AS rank_salary FROM employees WHERE rank_salary <= 3;`

Question 39

How do you count missing values in each column using Python & Pandas?

Answer 39

`df.isnull().sum()`

Question 40

How do you drop duplicate rows in Python & Pandas?

Answer 40

`df.drop_duplicates(inplace=True)`

Question 41

How do you get the top 5 most common values in a column using Python & Pandas?

Answer 41

`df['category'].value_counts().head(5)`

Question 42

How do you group by and compute the mean in Python & Pandas?

Answer 42

`df.groupby('category')['sales'].mean()`

Question 43

How do you apply a function to clean a text column in Python & Pandas?

Answer 43

`df['clean_text'] = df['text'].apply(lambda x: x.lower().strip())`

Question 44

How do you merge two DataFrames in Python & Pandas?

Answer 44

`merged_df = pd.merge(df1, df2, on='customer_id', how='left')`

Question 45

What is variance?

Answer 45

Variance measures how spread out data is from the mean.

Question 46

What is the difference between mean and median?

Answer 46

Mean is the average, median is the middle value.

Question 47

How do you detect outliers?

Answer 47

Using IQR, Z-scores, or visualization (boxplot, histogram).

Question 48

Explain Bayes’ Theorem

Answer 48

It calculates conditional probability: P(A|B) = P(B|A) * P(A) / P(B).

Question 49

When should you use T-test vs. Chi-square vs. ANOVA?

Answer 49

T-test compares two means, Chi-square tests categorical data, ANOVA compares multiple means.

Question 50

What’s a p-value?

Answer 50

The probability of observing results at least as extreme as the null hypothesis.

Question 51

What is multicollinearity, and how can you detect it?

Answer 51

High correlation between features, detected using VIF scores.

Question 52

How do you handle imbalanced datasets?

Answer 52

Use oversampling, undersampling, SMOTE, or adjust class weights.

Question 53

What is the difference between Linear & Logistic Regression?

Answer 53

Linear predicts continuous values, logistic predicts probabilities.

Question 54

What is overfitting and how can you prevent it?

Answer 54

Overfitting happens when a model learns noise; prevent it using regularization, dropout, or more data.

Question 55

Explain Precision, Recall, F1-score

Answer 55

Precision = TP / (TP + FP), Recall = TP / (TP + FN), F1 balances both.

Question 56

What is the difference between L1 (Lasso) & L2 (Ridge) Regularization?

Answer 56

L1 shrinks weights to zero, L2 penalizes large weights.

Question 57

What’s the use of Cross-Validation?

Answer 57

Prevents overfitting by evaluating models on different subsets of data.

Question 58

Why is Feature Scaling important?

Answer 58

Ensures equal weight for all features, improves gradient descent.

Question 59

How does PCA work for dimensionality reduction?

Answer 59

It projects data onto new axes to capture maximum variance.

Question 60

Explain Random Forest vs. Gradient Boosting

Answer 60

Random Forest reduces variance, boosting corrects model errors iteratively.

Question 61

How would you detect fake reviews?

Answer 61

Use NLP, sentiment analysis, and user behavior patterns.

Question 62

How would you improve a recommendation system?

Answer 62

Hybrid filtering, better embeddings, A/B testing models.

Question 63

How do you analyze a drop in user engagement?

Answer 63

Check retention, session data, A/B tests, and feature changes.

Question 64

How would you optimize a customer churn model?

Answer 64

Use classification models, analyze key churn indicators, improve retention strategies.

Question 65

Describe a time you worked with messy data.

Answer 65

Clearing Messy Data: 1. Remove duplicate rows 2. Handle missing values 3. Convert data types (put dates in datetime format) 4. Standardize text data (str.lower().strip()) 5. Handle outliers 6. Normalize or scale features (min max scale or z-score normalization) 7. Fix inconsistent categories (instead of having some labels as NY and some as New York for example) 8. Convert categories to numbers (label encoding or one hot encoding) 9. Create derived variables (new variables based on OG variables to help improve model performance, ie parts of a date or age from birthdate) 10. Verify data consistency (ensure all data adheres to same rules, ie consistent formats, currency, no impossible ages or dates, the ranges of data are realistic)

Question 66

Describe a challenging data project and how you solved it.

Answer 66

Discuss the problem, your approach, and impact.

Question 67

How do you communicate insights to non-technical stakeholders?

Answer 67

Use visuals, storytelling, and focus on business impact.

Question 68

What does the acronym STAR stand for in the STAR interview method?

Answer 68

Situation, Task, Action, Result