AI-900 Question 102
Select 3A company is building an image recognition system to classify different types of animals. They decide to use a deep learning model for this task. Which of the following are features of deep learning techniques that make them suitable for this scenario?
- A
Deep learning models can automatically extract features from raw data without requiring manual feature engineering.
- B
Deep learning requires significantly less labeled data compared to traditional machine learning techniques.
- C
Deep learning can process unstructured data types such as images, audio, and text effectively.
- D
Deep learning models rely heavily on shallow architectures with a small number of layers.
- E
Deep learning techniques are highly scalable and can leverage large amounts of data for improved accuracy.
Show answer and explanation
Correct answers: A, C, E
Explanation
Deep learning is particularly suited for tasks like image recognition due to its ability to automatically extract features, handle unstructured data, and scale effectively with large datasets. These features make it a powerful tool for building advanced AI systems.
- A. Correct.
Correct. One of the key features of deep learning is its ability to automatically extract features from raw data, such as images, without the need for manual intervention.
- B. Incorrect.
Incorrect. Deep learning typically requires a large amount of labeled data to achieve high performance, especially for tasks like image recognition.
- C. Correct.
Correct. Deep learning excels at processing unstructured data types, including images, audio, and text, which makes it ideal for tasks like image classification.
- D. Incorrect.
Incorrect. Deep learning uses deep architectures with multiple layers (hence the term 'deep') to learn complex patterns in data, not shallow architectures.
- E. Correct.
Correct. Deep learning techniques are highly scalable and perform well with large datasets, improving accuracy as the amount of training data increases.