AI /ML By INDOMITABLE LIMITEDOctober 12, 2025October 24, 2025 Welcome to your AI /ML 1. AI/ML Which of the following is an example of supervised learning? WhichK-Means Clustering Apriori Algorithm Linear Regression d) PCA None 2. AI/ML In machine learning, what is a ‘feature’? A data attribute used as input The final prediction A type of loss function The algorithm used None 3. AI/ML Which of the following is a classification algorithm? K-Means Apriori Decision Tree PCA None 4. AI/ML What is the main goal of supervised learning? Find patterns without labels Predict output using labeled data Reduce data size Perform dimensionality reduction None 5. AI/ML Which library is widely used for ML in Python? Flask NumPy BeautifulSoup TensorFlow None 6. AI/ML Which metric is commonly used for regression? Mean Squared Error Precision Accuracy Recall None 7. AI/ML Which of these is an unsupervised algorithm? K-Means Random Forest Naive Bayes Logistic Regression None 8. AI/ML What is overfitting? Model fits too well on training data but fails on new data Model performs well on all data Model uses too many outputs Model has too few features None 9. AI/ML Which of the following is NOT a type of machine learning? Predictive Analytics Unsupervised Supervised Reinforcement None 10. AI/ML Which activation function is most commonly used in hidden layers? ReLU Sigmoid Tanh Softmax None 11. AI/ML Which algorithm is best suited for text classification? K-Means Naive Bayes Linear Regression KNN None 12. What does PCA stand for? Partial Cluster Analysis Principal Correlation Approach Principal Component Analysis Predictive Component Algorithm None 13. AI/ML Gradient Descent is used to: Find missing values Minimize the cost function Maximize loss Normalize data None 14. AI/ML Which of these models is an ensemble method? Naive Bayes Random Forest K-Means Linear Regression None 15. AI/ML In neural networks, what is a ‘bias’? Error between prediction and truth Type of regularization Weight assigned to output Constant added to activation function None 16. AI/ML Which evaluation metric is used for imbalanced datasets? Precision-Recall or F1 Score RMSE Accuracy R² Score None 17. AI/ML Which optimizer is an improvement over traditional Gradient Descent? Softmax Sigmoid Min-Max Adam None 18. AI/ML Which of these techniques can help reduce overfitting? Removing validation data Dropout Regularization Adding more layers Increasing learning rate None 19. AI/ML What is the main difference between classification and regression? Classification uses unsupervised data Regression uses trees, classification doesn’t Both predict continuous values Classification predicts categories, regression predicts continuous values None 20. AI/ML What does the “kernel trick” refer to in SVMs? Reducing number of features Avoiding gradient descent Initializing weights Transforming data into higher dimensions None 21. AI/ML What is vanishing gradient problem? Gradients become too small during backpropagation Model doesn’t learn due to high learning rate Loss never decreases Gradients explode in size None 22. AI/ML What does the softmax function output? Probabilities that sum to 1 Raw scores Gradients Binary values None 23. AI/ML What is the purpose of the learning rate in optimization? Measures accuracy Controls batch size Defines model complexity Determines size of steps during gradient updates None 24. AI/ML What is the purpose of a confusion matrix? Detect overfitting Measure loss Visualize classification performance Show regression line None 25. AI/ML Which of the following best describes Reinforcement Learning? Learning from unsupervised clusters Learning from labeled data Learning without any data Learning through rewards and penalties None 26. None
