AI Engineer By INDOMITABLE LIMITEDOctober 14, 2025October 24, 2025 Welcome to your AI Engineer AI engineer 1. Which algorithm is used to minimize the cost function in neural networks? Naïve Bayes K-Means Gradient Descent Random Forest None AI engineer 2. What is the main purpose of an activation function in a neural network? Increase training speed Introduce non-linearity Normalize input data Reduce overfitting None AI engineer 3. Which of the following is a supervised learning algorithm? K-Means DBSCAN Linear Regression PCA None AI engineer What is the function of a loss function in machine learning? It stores weights It normalizes inputs It visualizes data It measures model error None AI engineer In deep learning, what does ReLU stand for? Random Linear Unit Recurrent Linear Unit Regularized Linear Update Rectified Linear Unit None AI engineer Which layer in a neural network is responsible for feature extraction? Output layer Input layer Normalization layer Hidden layers None AI engineer What is overfitting in AI models? Model stops training early Model performs poorly on training data Model performs well on test data only Model performs well on training but poorly on test data None AI engineer Which of these is an example of an unsupervised learning problem? Clustering Customer Data Sentiment Analysis Spam Detection Stock Price Prediction None AI engineer Which of the following metrics is used for classification problems? Mean Absolute Error Mean Squared Error R² Score Accuracy None AI engineer What does CNN stand for in deep learning? Central Neural Network Complex Neural Node Convolutional Neural Network Clustered Neural Network None AI engineer Which algorithm is used for dimensionality reduction? Logistic Regression Decision Tree KNN PCA None AI engineer What does CNN stand for in deep learning? Convolutional Neural Network Complex Neural Node Central Neural Network Clustered Neural Network None AI engineer What does dropout do in neural networks? Removes features permanently Normalizes data Increases learning rate Decreases overfitting by randomly disabling neurons None AI engineer What is a hyperparameter? A model output A loss value A learned weight A parameter set before training None AI engineer Which type of neural network is best suited for sequential data? GAN SVM RNN CNN None AI engineer GAN stands for what? Graph Attention Network Generative Adversarial Network Gradient Applied Network General Adversarial Network None AI engineer What is backpropagation used for? Updating model parameters Cleaning data Storing gradients Initializing weights None AI engineer Which loss function is commonly used for classification? Cross-Entropy Loss Mean Squared Error Hinge Loss Smooth L1 Loss None AI engineer What is the role of the softmax function? Reduces bias Converts outputs to probabilities Normalizes inputs Increases learning rate None AI engineer Which library is most commonly used for deep learning in Python? TensorFlow Matplotlib NumPy Pandas None AI engineer What is a confusion matrix used for? Feature selection Weight initialization Displaying correlations Evaluating classification performance None AI engineer In NLP, what does tokenization mean? Splitting text into words or subwords Compressing text Removing stopwords Encoding data None AI engineer Which optimizer adapts learning rate for each parameter? SGD Adam RMSprop MiniBatch None AI engineer What is the purpose of regularization? Increase model complexity Prevent overfitting Speed up training Normalize data None AI engineer In reinforcement learning, what is a policy? A mapping from states to actions An environment A reward function A discount factor None
