DY0-001: CompTIA DataAIDemo

Question

A data scientist wants to model a complex phenomenon using a large dataset containing categorical, discrete, and continuous variables. Following exploratory data analysis, the data scientist is reasonably confident that no linear relationship exists between the predictors and the target. Although the phenomenon is complex, the data scientist wants to retain the highest possible degree of interpretability in the final model. Which of the following algorithms best satisfies this objective?

A Artificial neural network
B Decision tree
C Multiple linear regression
D Random forest

Explanation

Decision trees model nonlinear relationships through successive feature-based split rules and can accommodate different variable types. A single tree is inherently interpretable because its predictions can be traced through explicit decision rules; random forests and neural networks are less transparent, while multiple linear regression assumes a linear relationship.

Learn more

Decision Trees — scikit-learn documentation

DY0-001: CompTIA DataAIDemo

AstroTutor

Notes

Readiness

QuestionQ58

Community Discussion