Which statement best describes the relationship between AI, machine learning (ML), and deep learning (DL)?
Ai Overview
Question · Easy
Overfitting: I memorized the practice test
Short
Which of the following is the best example of narrow (weak) AI rather than general (strong) AI?
Which event is widely cited as marking the formal birth of artificial intelligence as a research field?
Tech Exam Prep Podcast
Listen on Spotify
Podcast
Which of the following is NOT a core application domain of modern AI as covered in HCIA-AI?
Which type of machine learning task uses labeled data to predict a continuous numerical output, such as a house price?
Machine Learning Fundamentals
Which algorithm is most commonly used as the baseline for binary classification problems and outputs class probabilities via the sigmoid function?
Machine Learning Algorithms
Which of the following is an unsupervised learning algorithm?
What is the primary purpose of dimensionality reduction techniques such as PCA?
Question · Medium
In reinforcement learning, what does the agent receive from the environment after taking an action?
Which evaluation metric is most appropriate for a highly imbalanced binary classification problem, such as fraud detection where only 1% of cases are positive?
Model Evaluation
A confusion matrix shows TP=80, FP=20, FN=10, TN=890. What is the precision?
What does a ROC AUC of 0.5 indicate about a binary classifier?
Which technique is NOT a standard way to combat overfitting in machine learning?
What does L1 regularization tend to produce in a linear model that L2 does not?
Which describes underfitting?
Which technique splits the training data into K subsets and rotates which subset is used for validation, providing a more robust estimate of model performance?
Which activation function outputs zero for negative inputs and the input itself for non-negative inputs?
Deep Learning Fundamentals
Which loss function is standard for multi-class classification with a softmax output layer?
Which optimizer adapts the learning rate per parameter using running averages of gradients and squared gradients?
Which neural network architecture is specifically designed to exploit spatial structure in images using shared weights and local receptive fields?
Neural Network Types
What is the main benefit of residual (skip) connections introduced by ResNet?
Which RNN variant uses input, forget, and output gates to better model long-range dependencies in sequences?
Which mechanism is the foundation of the Transformer architecture and large language models like Pangu and GPT?
Which Huawei deep learning framework is the primary open-source framework promoted by Huawei for AI development?
Huawei Ai Ecosystem
Which Huawei one-stop AI development platform covers data management, model training, deployment, and inference?
Which Huawei Ascend AI processor is optimized primarily for inference workloads at the edge and in low-power scenarios?
Which Huawei Ascend processor is the high-performance chip designed primarily for AI training in data centers?
What does CANN stand for in the Huawei AI software stack?
Which Atlas product is typically used as a high-density AI training server powered by Ascend 910 processors?
Which Huawei Cloud AI service would you most likely use to convert spoken audio into written text?
Huawei Cloud Services
Which Huawei Cloud service helps users automatically build machine learning models with little or no coding?
Which Huawei Cloud AI service extracts structured text from documents such as ID cards, invoices, and license plates?
Which is a primary smart city use case enabled by AI on Huawei platforms?
Huawei Applications
Which AI ethics principle most directly addresses the concern that training data may underrepresent certain groups, leading to unfair predictions?
Ai Ethics
Which European regulation primarily governs personal data protection and is frequently referenced in AI ethics discussions?
Which is a defining characteristic of a foundation model (such as Pangu, GPT, or Llama)?
Generative Ai
Which decision tree splitting criterion measures the impurity of a node based on the probability distribution of classes, with 0 representing perfect purity?
What is the main advantage of a random forest over a single decision tree?
Which algorithm finds the maximum-margin hyperplane that separates classes and can use the kernel trick to handle non-linear boundaries?
Which characteristic distinguishes DBSCAN from K-means clustering?
Which gradient boosting framework is widely used for tabular data in Kaggle competitions and production ML systems?
What is the key idea behind the K-nearest neighbors (KNN) algorithm?
Which assumption does naive Bayes make that is rarely true in practice but still allows it to perform surprisingly well?
Why is the typical machine learning workflow split into training, validation, and test sets?
Which metric pair would be reported together to evaluate a regression model trained with MindSpore?
Which scenario is best addressed by recommendation systems, one of the four core AI application domains?
Loading more…