AWS Certified AI Practitioner Practice Exam

Which ML algorithm is suitable for classifying human genes into categories and documenting the model's decision process?

1. Decision trees

2. Linear regression

3. Logistic regression

4. Neural networks

The correct answer is: Decision trees

Decision trees are suitable for classifying human genes into categories because they work by splitting the data into subsets based on the values of the features, ultimately leading to a decision based on specific criteria. Each split is made to increase the purity of the categories, which makes decision trees an intuitive and interpretable algorithm. One of the significant advantages of decision trees is their ability to document the model's decision-making process clearly and understandably. The structure of a decision tree visually represents the path taken for classification, allowing researchers and practitioners to see how specific genes are categorized based on the features used, making it easier to trace back the reasoning behind each classification. While neural networks can effectively classify complex data, they often operate as "black boxes," making interpreting the decision process challenging. Similarly, linear and logistic regression are primarily used for predicting numerical outcomes or binary classifications based on relationships between features, but they lack the ability to categorize in a manner similar to decision trees. Moreover, while logistic regression could be used for binary classifications, it does not provide the same level of interpretability as decision trees, particularly for multi-class problems that might be encountered in gene classification.