Feature engineering basics in Pandas - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is feature engineering in data science?

Feature engineering is the process of creating new input variables or modifying existing ones to improve a machine learning model's performance.

beginner

Why is feature engineering important?

It helps models learn better by providing more meaningful or clearer information, often leading to improved accuracy and insights.

beginner

Name three common feature engineering techniques.

1. Creating new features by combining existing ones (e.g., ratios)<br>2. Encoding categorical variables (e.g., one-hot encoding)<br>3. Scaling or normalizing numerical features

beginner

How can you create a new feature in pandas?

You can create a new column by performing operations on existing columns, for example: df['new_feature'] = df['A'] + df['B'].

beginner

What is one-hot encoding and when do you use it?

One-hot encoding converts categorical variables into multiple binary columns (0 or 1) to help models understand categories as numbers without implying order.

Which of the following is NOT a feature engineering technique?

A Creating new features from existing data

B Removing missing values

C Encoding categorical variables

D Scaling numerical features

Removing missing values is data cleaning, not feature engineering.

In pandas, how do you create a new feature 'total' by adding columns 'A' and 'B'?

A df['total'] = df['A'] + df['B']

B df.total = df.A + df.B

C df.add('total', df['A'], df['B'])

D df['total'] = df['A'] * df['B']

Option A correctly creates a new column by adding two columns.

What does one-hot encoding do?

A Converts numerical data into categories

B Scales features between 0 and 1

C Turns categorical variables into binary columns

D Removes duplicate rows

One-hot encoding creates binary columns for each category.

Why might you create a new feature like 'age_group' from 'age'?

A To group continuous data into meaningful categories

B To remove outliers

C To reduce the number of missing values

D To encode text data

Grouping continuous data into categories can help models capture patterns better.

Which pandas function helps to scale numerical features?

A df.drop_duplicates()

B df.fillna()

C pd.get_dummies()

D sklearn.preprocessing.StandardScaler()

StandardScaler from sklearn is used to scale numerical features.

Explain what feature engineering is and why it matters in simple terms.

Think about how changing or adding data columns can help a model learn better.

Describe how you would create a new feature in a pandas DataFrame using existing columns.

Remember the syntax for adding a new column in pandas.