Тест: Introduction to AI and Machine Learning — 9 въпроса

1. In what year did Alan Turing propose the Turing Test as a measure of AI's human-like ability?

A. 1970

B. 1945

C. 1950

D. 1960

Answer: (C)

Обяснение

Alan Turing proposed the Turing Test in 1950 as a benchmark for evaluating a machine’s ability to exhibit human-like intelligence. The other years are plausible but incorrect, serving as distractors.

2. Who proposed the Turing Test as a benchmark for evaluating artificial intelligence?

A. Claude Shannon

B. John McCarthy

C. Marvin Minsky

D. Alan Turing

Answer: (D)

Обяснение

Alan Turing proposed the Turing Test in 1950 as a way to evaluate a machine's ability to exhibit human-like intelligence. The other figures listed contributed significantly to AI and computer science but are not credited with proposing the Turing Test.

3. What is a key characteristic that distinguishes supervised learning from other machine learning approaches?

A. It relies on unlabeled data to discover hidden patterns.

B. It automatically extracts features from raw data without labels.

C. It involves reinforcement signals like rewards or penalties.

D. It uses labeled datasets to train models for predictions.

Answer: (D)

Обяснение

Supervised learning is characterized by training models on labeled datasets, where each input has a known output, enabling the model to learn the mapping from inputs to outputs for future predictions. This distinguishes it from unsupervised learning, which does not use labels, and reinforcement learning, which relies on reward signals.

4. When was the concept of Unsupervised Learning first established as a distinct category within machine learning?

A. In the early 1950s during initial AI research

B. In the 2000s with the advent of big data

C. In the 1960s with the development of clustering algorithms

D. In the 1980s during the rise of neural networks

Answer: (C)

Обяснение

Unsupervised Learning as a distinct category gained recognition in the 1960s, notably with the development of clustering algorithms like K-means, which marked its formal establishment within machine learning research.

5. What is essential when applying deep learning models to real-world problems?

A. Collecting large datasets and utilizing powerful computational hardware

B. Applying shallow neural networks with fewer layers for better performance

C. Relying solely on manual feature engineering instead of automatic learning

D. Using only small, curated datasets to prevent overfitting

Answer: (A)

Обяснение

Deep learning models require large datasets and significant computational resources, such as GPUs or TPUs, to effectively learn complex hierarchical patterns from raw data. This is emphasized in the context as a key factor for success in practical applications. The other options are incorrect: small datasets limit deep learning effectiveness; manual feature engineering is less necessary in deep learning; shallow networks do not leverage the full potential of deep neural architectures.

6. What is the primary role of neural networks in machine learning?

A. To learn complex data patterns and relationships

B. To automatically generate new data samples

C. To store large amounts of data efficiently

D. To simulate human decision-making processes

Answer: (A)

Обяснение

Neural networks are designed to learn complex patterns and relationships in data through training, enabling tasks like classification, prediction, and recognition. They do not primarily serve as data storage, decision-making alone, or data generation, although they can be involved indirectly in these processes.

7. What was a direct consequence of Alan Turing proposing the Turing Test in 1950?

A. It caused the discontinuation of rule-based AI systems.

B. It became a foundational benchmark for assessing AI's human-like capabilities.

C. It immediately led to the creation of general AI systems.

D. It resulted in AI being considered a purely philosophical field.

Answer: (B)

Обяснение

The proposal of the Turing Test by Alan Turing in 1950 became a foundational concept and a benchmark for evaluating whether machines can exhibit human-like intelligence, influencing how AI systems are assessed and debated.

8. How do neural networks differ from the broader field of computer vision?

A. Neural networks are unrelated to computer vision and are used for natural language processing.

B. Neural networks are the only method used in computer vision for image recognition tasks.

C. Neural networks are a subset of techniques used specifically for analyzing visual data within computer vision.

D. Neural networks are a hardware technology, whereas computer vision is a software field.

Answer: (C)

Обяснение

Neural networks are a subset of techniques used within computer vision, particularly for tasks like image recognition and object detection. They are not the only method, but a powerful and widely used one. The other options are incorrect because neural networks are not the sole method, are related to computer vision, and are not hardware technologies.

9. What are AI applications?

A. The deployment of AI systems to solve real-world problems across various industries

B. The theoretical study of artificial intelligence algorithms and models

C. The development of AI hardware components for faster processing

D. The programming of machines to perform specific tasks without learning

Answer: (A)

Обяснение

AI applications refer to the practical deployment of AI technologies to address real-world problems in industries such as healthcare, finance, and autonomous systems. They involve using AI systems to improve efficiency, decision-making, and automation, distinguishing them from theoretical or purely technical aspects of AI development.