DEVELOPING AN ARTIFICIAL INTELLIGENCE CURRICULUM TU4

The steps to follow to learn a model are:

• Prepare the data, train the model and use the model

• Prepare the data, test the model and use the model

• Prepare the data, train the model, test the model and use the model

The property of intelligent systems so they are able of generalizing from experience is:

• Perception

• Reasoning

• Learning

The value that provides the probability of recognizing an object is called:

• Information value

• Confidence value

• Error rate

The quality of the images in the dataset for training depends on:

• Their size and format

• The variety in orientations and the contrast from the background

• The number of objects included in them

The developed game does not use any sensor, and therefore, does not work the perception

topic.

• True

• False

The models learned in this TU are able to detect if an object is not one of the trained ones.

• No, they always consider it one of the trained objects.

• Yes, they can detect it.

• Depends on whether the object looks like one of the trained ones.

The model created not only learns when it is trained, but also learns from its mistakes.

• True

• False