Data Science with AI Interview Preparation Guide : 240 Q&As

Q76. What is Machine Learning?

Machine Learning allows systems to learn patterns from data and make predictions without being explicitly programmed for every rule.

Q77. What are the main types of Machine Learning?

The main types are:

• Supervised Learning

• Unsupervised Learning

• Reinforcement Learning

Q78. What is Supervised Learning?

Supervised Learning uses labeled data.

Example:

Input: Customer details
Output: Churn or Not Churn

Q79. What is Unsupervised Learning?

Unsupervised Learning works with unlabeled data.

It is commonly used for clustering and pattern discovery.

Q80. What is Reinforcement Learning?

Reinforcement Learning learns through rewards and penalties based on actions.

Q81. What is Regression?

Regression predicts a continuous value.

Example:

House Price Prediction

Q82. What is Classification?

Classification predicts a category or class.

Example:

Spam or Not Spam

Q83. Regression vs Classification?

Regression: Predicts numbers.

Classification: Predicts categories.

Q84. What is Linear Regression?

Linear Regression models the relationship between input variables and a continuous target.

Q85. What is Logistic Regression?

Logistic Regression is mainly used for classification problems.

It predicts probabilities for classes.

Q86. What is a Decision Tree?

A Decision Tree makes predictions using a sequence of conditions.

It is easy to understand but can overfit.

Q87. What is Random Forest?

Random Forest combines many decision trees and uses their combined output.

It usually reduces overfitting compared with one tree.

Q88. What is K-Nearest Neighbors?

KNN predicts based on the nearest data points.

It works better when features are properly scaled.

Q89. What is Naive Bayes?

Naive Bayes is a probabilistic algorithm often used for text classification.

It assumes features are conditionally independent.

Q90. What is Support Vector Machine?

SVM finds a boundary that separates classes with maximum margin.

It works well for some high-dimensional problems.

Q91. What is clustering?

Clustering groups similar data points together without predefined labels.

Q92. What is K-Means?

K-Means divides data into K clusters based on similarity.

Q93. How do you choose K in K-Means?

Common methods include:

• Elbow Method

• Silhouette Score

• Business understanding

Q94. What is overfitting?

Overfitting happens when a model learns training data too closely and performs poorly on new data.

Q95. What is underfitting?

Underfitting happens when the model is too simple to learn the important patterns.

Q96. How can overfitting be reduced?

Common methods include:

• More data

• Regularization

• Simpler model

• Cross-validation

• Pruning

• Dropout for neural networks

Q97. What is train-test split?

The dataset is divided into:

Training data: Used to learn.

Test data: Used to evaluate final performance.

Q98. What is validation data?

Validation data is used to compare models and tune settings before final testing.

Q99. What is cross-validation?

Cross-validation trains and evaluates a model on different data splits.

It gives a more reliable estimate of model performance.

Q100. What is a feature?

A feature is an input variable used by the model.

Example:

Age, Salary, Purchase Count

Q101. What is a target variable?

The target is the value the model tries to predict.

Example:

Customer Churn

Q102. What is feature scaling?

Feature scaling brings numerical features to comparable ranges.

It is useful for algorithms such as KNN and SVM.

Q103. What is normalization?

Normalization usually scales values into a fixed range such as 0 to 1.

Q104. What is standardization?

Standardization transforms data around:

Mean = 0

Standard Deviation = 1

Q105. What is hyperparameter tuning?

Hyperparameter tuning finds better model settings.

Examples:

• Tree depth

• Number of trees

• Learning rate

Q106. What is Grid Search?

Grid Search tests predefined combinations of hyperparameters.

Q107. What is Random Search?

Random Search tests randomly selected hyperparameter combinations.

It can be faster when many parameters exist.

Q108. What is ensemble learning?

Ensemble learning combines multiple models to improve prediction quality.

Examples:

Random Forest, Bagging, Boosting

Q109. What is Bagging?

Bagging trains multiple models independently and combines their results.

It mainly helps reduce variance.

Q110. What is Boosting?

Boosting trains models sequentially, where later models focus more on previous errors.

Examples:

AdaBoost, Gradient Boosting, XGBoost