AI/ML Screening TestBy mfh.officials@gmail.com / January 7, 2025 Artificial Intelligence (AI) and Machine Learning (ML) 1 / 18 Which type of machine learning algorithm is used to group data points into distinct clusters? Dimensionality Reduction Regression Classification Clustering 2 / 18 What is "bias" in machine learning? The error due to incorrect assumptions made by the model The error due to model complexity The error due to random fluctuations in the dataset The error due to lack of data 3 / 18 Which of the following algorithms is typically used for outlier detection in continuous data? Naive Bayes Decision Trees DBSCAN K-Means 4 / 18 In deep learning, which of the following techniques is used to prevent the model from overfitting? All of the above Early Stopping L2 Regularization Dropout 5 / 18 Which algorithm is best suited for detecting anomalies in data? K-Means Clustering One-Class SVM Decision Trees Naive Bayes 6 / 18 Which technique is used to evaluate the performance of a regression model? Mean Squared Error (MSE) ROC Curve Accuracy Precision 7 / 18 Which technique can be used to prevent overfitting in deep learning models? Cross-validation All of the above Early stopping Regularization (L1/L2) 8 / 18 What does the "softmax" activation function do? Maps any value between 0 and 1 Converts the output of a network to a probability distribution Introduces non-linearity to the network Prevents overfitting in the model 9 / 18 Which of the following is true for the concept of "Gradient Descent"? It works by choosing random weights for model parameters. It always converges to the global minimum. It is used to optimize a function by finding the local minimum. It cannot be used for deep learning models. 10 / 18 Which of the following is NOT a type of Machine Learning? Unsupervised Learning Evolutionary Learning Reinforcement Learning Supervised Learning 11 / 18 Which type of neural network is specifically designed for spatial data, such as images? Multilayer Perceptrons (MLP) Generative Adversarial Networks (GAN) Convolutional Neural Networks (CNN) Recurrent Neural Networks (RNN) 12 / 18 Which of the following is true about decision trees? They are sensitive to overfitting only on very small datasets. They split data based on feature values using a tree-like structure. They are not interpretable. They can only be used for classification tasks. 13 / 18 Which of the following models uses the concept of a "kernel" to map input data into a higher-dimensional space? Support Vector Machines (SVM) Naive Bayes Decision Trees Logistic Regression 14 / 18 Which neural network architecture is designed to work with sequential data such as time-series or text? Convolutional Neural Networks (CNN) Deep Belief Networks Recurrent Neural Networks (RNN) Fully Connected Networks 15 / 18 Which of the following techniques is used to improve the performance of an underfitting model? Cross-validation Regularization Increase model complexity Pruning 16 / 18 What is the purpose of "feature scaling" in machine learning? To make sure all features are on the same scale to improve model performance To increase the interpretability of the model To reduce the number of features To eliminate missing values 17 / 18 Which of the following is a characteristic of supervised learning? The model uses feedback from the environment. The model learns from labeled data. The model learns to classify data into distinct categories without labels. The model learns patterns without labeled data. 18 / 18 Which of the following measures the proportion of false positives in a classification problem? F1 Score Precision False Positive Rate (FPR) Recall Your score isThe average score is 48%
By
/ January 7, 2025 By
/ January 7, 2025 By
/ January 7, 2025 By
/ January 7, 2025 By
/ January 7, 2025 By
/ January 7, 2025