Feature Selection and Feature Engineering

The video tutorial covers the concepts of feature selection and feature engineering in machine learning. It explains the difference between features and labels, the importance of selecting relevant features, and the benefits of reducing redundancy. The tutorial also introduces feature engineering as a method to create new features to enhance predictive power. Practical examples and questions are provided to reinforce understanding.

What is the term used for the column that an algorithm aims to predict?

• Attribute

• Variable

• Label

• Feature

Which of the following is NOT a reason to perform feature selection?

• To increase the number of features

• To decrease training time

• To improve model accuracy

• To reduce overfitting

What is the principle of 'minimum redundancy' in feature selection?

• Maximizing the number of features

• Ensuring all features are unique

• Removing features that are similar or duplicate

• Using as many features as possible

How does feature selection help in reducing overfitting?

• By using only relevant features

• By increasing the number of features

• By simplifying the algorithm

• By adding more data

What is feature engineering?

• The process of labeling data

• The process of removing irrelevant features

• The process of creating new features from raw data

• The process of selecting existing features

Which of the following is an example of feature engineering?

• Splitting a date into day, month, and year

• Selecting only numerical columns

• Dropping columns with missing values

• Removing duplicate columns

What is the main goal of feature engineering?

• To reduce the dataset size

• To enhance the predictive power of a model

• To simplify the data collection process

• To automate the feature selection process