Data Science and Machine Learning (Theory and Projects) A to Z - Introduction to Machine Learning: Machine Learning Mode

The video tutorial introduces multi-target models, which predict multiple outputs or vectors instead of a single target. It explains that these models are common in regression tasks but can also be used in classification to predict probabilities for different classes. The tutorial provides examples of input-output pairs and discusses the underlying theory, emphasizing that the principles of multi-target modeling are similar to those of single-target models.

What is a key characteristic of multi-target models?

• They require equal-sized input and output vectors.

• They only work with classification tasks.

• They predict a vector of outputs.

• They predict a single output.

In which type of tasks are multi-target models more commonly used?

• Dimensionality reduction tasks

• Clustering tasks

• Regression tasks

• Classification tasks

What is an example of an output in a multi-target regression model?

• A vector of real numbers

• A vector of probabilities

• A single probability score

• A single real number

How does the training data for multi-target models differ from single-target models?

• The input and output vectors must be the same size.

• The input vector is always larger than the output vector.

• The number of coordinates in input and output vectors can differ.

• The output vector is always larger than the input vector.

What remains consistent in multi-target models compared to single-target models?

• The type of tasks they are used for

• The underlying theory

• The size of input vectors

• The number of outputs