Top Fresher Jobs in AI, ML, Data Science – 17.10.2025

Top 25 Data Science Interview Questions (with Answers)

1. What is Data Science?

Answer:
Data Science is a multidisciplinary field that uses statistics, computer science, and domain knowledge to extract meaningful insights from structured and unstructured data. It combines techniques from machine learning, data analysis, and data visualization to support data-driven decisions.

2. What are the main steps in a Data Science project?

Answer:

1. Data Collection

2. Data Cleaning & Preprocessing

3. Exploratory Data Analysis (EDA)

4. Feature Engineering

5. Model Building

6. Model Evaluation

7. Deployment and Monitoring

3. What is the difference between supervised and unsupervised learning?

Answer:

• Supervised Learning: Uses labeled data (e.g., classification, regression).

• Unsupervised Learning: Uses unlabeled data to find hidden patterns (e.g., clustering, association).

4. What is overfitting? How can it be prevented?

Answer:
Overfitting occurs when a model performs well on training data but poorly on new data.
Prevention: Use cross-validation, regularization (L1/L2), pruning, or early stopping.

5. What is cross-validation?

Answer:
Cross-validation is a model validation technique that splits data into training and testing sets multiple times (e.g., k-fold CV) to ensure the model generalizes well.

6. What is the difference between variance and bias?

Answer:

• Bias: Error from oversimplifying the model (underfitting).

• Variance: Error from sensitivity to small data changes (overfitting).
  A good model maintains a bias-variance tradeoff.

7. Explain precision and recall.

Answer:

• Precision: Out of all predicted positives, how many are correct.

• Recall: Out of all actual positives, how many were correctly predicted.
  They are combined using the F1-score for balanced performance evaluation.

8. What is the difference between classification and regression?

Answer:

• Classification: Predicts discrete labels (e.g., spam or not spam).

• Regression: Predicts continuous values (e.g., house prices).

9. What are outliers, and how do you handle them?

Answer:
Outliers are data points significantly different from others.
Handling methods:

• Remove them using IQR or z-score.

• Transform data (log/box-cox).

• Cap values (Winsorization).

10. What is feature engineering?

Answer:
Feature engineering involves creating, transforming, or selecting variables (features) to improve model performance. It includes encoding categorical data, scaling, and deriving new features.

11. What is the difference between bagging and boosting?

Answer:

• Bagging: Builds independent models in parallel (e.g., Random Forest).

• Boosting: Builds models sequentially, focusing on errors (e.g., XGBoost, AdaBoost).

12. What is normalization vs. standardization?

Answer:

• Normalization: Scales data between 0 and 1.

• Standardization: Transforms data to have mean = 0 and standard deviation = 1.

13. What are common Python libraries used in Data Science?

Answer:

• Data Handling: pandas, NumPy

• Visualization: Matplotlib, Seaborn, Plotly

• Machine Learning: scikit-learn, TensorFlow, PyTorch

• Data Cleaning: re (regex), missingno

14. What is a confusion matrix?

Answer:
A confusion matrix shows actual vs. predicted classifications — with True Positives (TP), False Positives (FP), True Negatives (TN), and False Negatives (FN). It helps calculate precision, recall, and accuracy.

15. Explain the Central Limit Theorem (CLT).

Answer:
The CLT states that the sampling distribution of the sample mean approaches a normal distribution as the sample size increases, regardless of the population’s distribution.

16. What is p-value in hypothesis testing?

Answer:
The p-value indicates the probability of observing the given result under the null hypothesis.

• If p-value < 0.05, reject H₀ (statistically significant).

17. What is correlation vs. causation?

Answer:

• Correlation: Measures relationship strength between variables.

• Causation: Indicates one variable directly affects another.
  Correlation ≠ Causation.

18. What is the difference between SQL INNER JOIN and LEFT JOIN?

Answer:

• INNER JOIN: Returns only matching records from both tables.

• LEFT JOIN: Returns all records from the left table and matching ones from the right table.

19. What is regularization in machine learning?

Answer:
Regularization reduces overfitting by penalizing large coefficients.

• L1 (Lasso): Shrinks some weights to zero.

• L2 (Ridge): Reduces magnitude but keeps all weights.

20. What is the difference between PCA and LDA?

Answer:

• PCA (Principal Component Analysis): Unsupervised, focuses on variance.

• LDA (Linear Discriminant Analysis): Supervised, maximizes class separability.

21. What is the difference between mean, median, and mode?

Answer:

• Mean: Average of data points.

• Median: Middle value (robust to outliers).

• Mode: Most frequent value.

22. What are Type I and Type II errors?

Answer:

• Type I Error: False positive (rejecting a true null).

• Type II Error: False negative (failing to reject a false null).

23. What is a time series?

Answer:
A time series is a sequence of data points collected over time (e.g., stock prices).
Key components: trend, seasonality, cyclic, and noise.

24. What are some common distance metrics in clustering?

Answer:

• Euclidean Distance (most common)

• Manhattan Distance

• Cosine Similarity
  Used in algorithms like K-Means and KNN.

25. Describe a real-world data science project you would build.

Answer:
“I’d analyze customer churn data to predict which users are likely to leave. I’d clean the dataset, perform EDA, use logistic regression for prediction, and visualize key patterns to support retention strategies.”