AI-900 Question 105
Select 3A company is building a machine learning model to classify images of animals. They decide to use a deep learning technique to improve accuracy. Which of the following are key features of deep learning techniques that make them suitable for this task?
- A
Ability to automatically extract features from raw data
- B
Dependence on small datasets to achieve high accuracy
- C
Use of multiple layers of artificial neural networks
- D
Requirement for extensive feature engineering by domain experts
- E
Capability to process unstructured data like images and text
Show answer and explanation
Correct answers: A, C, E
Explanation
Deep learning techniques are particularly suited for tasks involving unstructured data due to their ability to learn hierarchical representations and automatically extract features. These models typically employ deep neural networks and excel when applied to large datasets, making them powerful tools for tasks like image classification.
- A. Correct.
Deep learning techniques can automatically extract features from raw data without manual intervention, a crucial advantage when working with complex data like images.
- B. Incorrect.
Deep learning typically requires large datasets to perform well, as more data enables the model to learn better representations.
- C. Correct.
Deep learning models use multiple layers of artificial neural networks (often referred to as 'deep' networks), which allow them to learn hierarchical representations of data.
- D. Incorrect.
Unlike traditional machine learning, deep learning reduces the need for extensive feature engineering, as it learns features directly from data.
- E. Correct.
Deep learning models excel at processing unstructured data such as images, text, and audio, making them ideal for tasks like image classification.