Creating interaction features in ML Python - Evaluation Workflow

Metrics & Evaluation - Creating interaction features

Which metric matters for creating interaction features and WHY

When we create interaction features, we want to see if they help the model learn better. The main metrics to check are validation accuracy or validation loss. These show if the model predicts better on new data, not just the training data.

If the interaction features improve these metrics, it means they add useful information. If not, they might just add noise or make the model too complex.

Confusion matrix example

Suppose we have a classification task. Here is a confusion matrix before and after adding interaction features:

Before interaction features:
| TP=40 | FP=10 |
| FN=15 | TN=35 |

After interaction features:
| TP=45 | FP=8  |
| FN=10 | TN=37 |
    

Adding interaction features increased true positives and true negatives, and reduced false negatives and false positives. This means better predictions.

Precision vs Recall tradeoff with interaction features

Interaction features can help balance precision and recall. For example:

• Precision measures how many predicted positives are correct.
• Recall measures how many actual positives are found.

If interaction features help the model find more true positives without adding many false positives, recall and precision both improve.

But if interaction features cause the model to predict too many positives, precision may drop even if recall rises.

We want to find interaction features that improve both or at least keep a good balance.

Good vs Bad metric values for interaction features

Good: After adding interaction features, validation accuracy increases, validation loss decreases, and precision and recall improve or stay stable.

Bad: Validation accuracy drops or stays the same, loss increases, or precision and recall get worse. This means interaction features are not helping.

Also watch for overfitting: if training accuracy improves but validation accuracy drops, interaction features might be too complex.

Common pitfalls when evaluating interaction features

• Overfitting: Interaction features can make the model too complex, fitting noise instead of real patterns.
• Data leakage: Creating interaction features using future or test data can give false good results.
• Ignoring validation: Only checking training metrics can mislead you about feature usefulness.
• Too many features: Adding many interaction features can slow training and confuse the model.

Self-check question

Your model has 98% accuracy but only 12% recall on fraud cases after adding interaction features. Is it good for production?

Answer: No. Even though accuracy is high, recall is very low. This means the model misses most fraud cases, which is bad for fraud detection. Interaction features did not help find fraud well.

Key Result

Interaction features should improve validation accuracy and balance precision and recall without causing overfitting.

Practice

1. What is the main purpose of creating interaction features in machine learning?

easy

A. To capture the combined effect of two or more features on the target

B. To reduce the number of features in the dataset

C. To normalize the features to a common scale

D. To remove irrelevant features automatically

Solution

1. Step 1: Understand interaction features

   Interaction features combine two or more features to capture their joint effect on the target variable.

2. Step 2: Compare options

   Only To capture the combined effect of two or more features on the target describes capturing combined effects, which is the purpose of interaction features.

3. Final Answer:

   To capture the combined effect of two or more features on the target -> Option A

4. Quick Check:

   Interaction features = combined effect [OK]

Hint: Interaction features capture combined effects of features [OK]

Common Mistakes:

• Confusing interaction features with feature scaling
• Thinking interaction features reduce feature count
• Assuming interaction features remove irrelevant features

2. Which of the following is the correct way to create an interaction feature between two numeric features x1 and x2 in Python?

easy

A. interaction = x1 * x2

B. interaction = x1 - x2

C. interaction = x1 / x2

D. interaction = x1 + x2

Solution

1. Step 1: Recall how interaction features are created

   Interaction features are typically created by multiplying numeric features to capture their joint effect.

2. Step 2: Check each option

   Only multiplication (x1 * x2) correctly creates an interaction feature.

3. Final Answer:

   interaction = x1 * x2 -> Option A

4. Quick Check:

   Interaction = multiply features [OK]

Hint: Multiply numeric features to create interaction features [OK]

Common Mistakes:

• Using addition instead of multiplication
• Using division or subtraction which do not capture interaction
• Confusing interaction with feature scaling

3. Given the code below, what will be the output of print(df['interaction'].tolist())?

import pandas as pd

df = pd.DataFrame({'x1': [1, 2, 3], 'x2': [4, 5, 6]})
df['interaction'] = df['x1'] * df['x2']
print(df['interaction'].tolist())

medium

A. [4, 5, 6]

B. [5, 7, 9]

C. [1, 2, 3]

D. [4, 10, 18]

Solution

1. Step 1: Calculate interaction feature values

   Multiply each pair: 1*4=4, 2*5=10, 3*6=18.

2. Step 2: Verify output list

   The list of interaction values is [4, 10, 18].

3. Final Answer:

   [4, 10, 18] -> Option D

4. Quick Check:

   Multiplying pairs = [4, 10, 18] [OK]

Hint: Multiply row-wise values for interaction feature list [OK]

Common Mistakes:

• Adding instead of multiplying features
• Confusing original features with interaction
• Misreading the DataFrame values

4. The following code attempts to create an interaction feature between two categorical features color and shape. What is the error?

import pandas as pd

df = pd.DataFrame({'color': ['red', 'blue'], 'shape': ['circle', 'square']})
df['interaction'] = df['color'] * df['shape']
print(df['interaction'])

medium

A. DataFrame columns must be numeric to create interaction

B. The DataFrame is missing a target column

C. You cannot multiply string columns directly; need encoding first

D. The print statement syntax is incorrect

Solution

1. Step 1: Understand data types for interaction

   Multiplying string columns causes an error because strings cannot be multiplied directly.

2. Step 2: Identify correct approach

   Categorical features must be encoded (e.g., one-hot or label encoding) before creating interaction features.

3. Final Answer:

   You cannot multiply string columns directly; need encoding first -> Option C

4. Quick Check:

   Multiply strings error = need encoding [OK]

Hint: Encode categorical features before multiplying [OK]

Common Mistakes:

• Trying to multiply raw string columns
• Ignoring data type requirements for interaction
• Assuming print syntax is wrong

5. You have two categorical features: Gender with values ['Male', 'Female'] and Smoker with values ['Yes', 'No']. How would you create an interaction feature to help a model learn their combined effect?

hard

A. Multiply the raw string columns directly

B. One-hot encode both features, then multiply corresponding columns

C. Add the string values together as new strings

D. Ignore interaction features for categorical data

Solution

1. Step 1: Encode categorical features

   Convert 'Gender' and 'Smoker' into one-hot encoded numeric columns.

2. Step 2: Create interaction features

   Multiply corresponding one-hot columns (e.g., Male*Yes) to capture combined effect.

3. Final Answer:

   One-hot encode both features, then multiply corresponding columns -> Option B

4. Quick Check:

   Encode then multiply categorical features [OK]

Hint: One-hot encode then multiply for categorical interaction [OK]

Common Mistakes:

• Trying to multiply raw strings
• Concatenating strings instead of encoding
• Skipping interaction features for categorical data