3.Question 3What term is used if your test data leaks into the training data?1 pointTest leakageTraining leakageData leakage Historical data leakage

Analyzing the Options

1. Test Leakage: This term suggests the test data has unintentionally influenced the training data, which does not precisely capture the broader concept.

2. Training Leakage: This term is less commonly used and could lead to confusion.

3. Data Leakage: This is the most recognized term in the context of machine learning, where information from the test dataset inadvertently influences the training phase, leading to overfitting and inaccurately high performance metrics.

4. Historical Data Leakage: This term is less common and typically refers to a specific scenario rather than the general issue of test data contaminating training data.

Conclusion

The correct term that describes the scenario where test data leaks into training data is Data Leakage. This term accurately depicts the issue where the integrity of the machine learning model may be compromised due to unintended access to future or "test" data during the training phase.

Final Answer

Data leakage is the term used if your test data leaks into the training data.