Evaluate the impact of an AI application used in the real world. (case study) : Working with Flower Images: Case Study -

The video tutorial discusses data augmentation, a technique used to enhance datasets by creating variations of existing data. It is particularly useful when dealing with insufficient data or outliers. Techniques such as flipping, rotation, cropping, and adding noise are employed to increase dataset size and improve model accuracy. The tutorial also explains the importance of data augmentation in reducing overfitting and handling lower quality data. Two types of transformations, linear and offline, are introduced, with examples illustrating their application in different contexts.

Why is data augmentation particularly useful when dealing with insufficient data?

• It helps in reducing the size of the dataset.

• It eliminates the need for data preprocessing.

• It increases the diversity and size of the dataset.

• It speeds up the training process.

Which of the following is NOT a technique used in data augmentation?

• Rotation

• Flipping

• Scaling

• Pooling

In the context of data augmentation, what does 'pooling' refer to?

• A method to enhance image resolution

• A technique to increase dataset size

• A process to increase invariance in data

• A way to reduce the number of data points

What is a primary reason for using data augmentation in model training?

• To reduce computational cost

• To handle overfitting

• To simplify the model architecture

• To increase the speed of training

Which statement is true about linear transformations in data augmentation?

• They always involve complex mathematical operations.

• They can include simple operations like flipping an image.

• They are only applicable to text data.

• They are not suitable for image data.

Why is it important to consider the context when applying transformations?

• To make the data look more appealing.

• To ensure transformations do not distort the data meaningfully.

• Because context does not matter in data augmentation.

• Because transformations are universally applicable.

What is the main difference between linear and offline transformations?

• Linear transformations are more complex than offline transformations.

• Linear transformations are applied in real-time, while offline transformations are pre-computed.

• Offline transformations are applied to text data only.

• Linear transformations require more computational resources.