Practice Questions

Explanation: A feasibility assessment evaluates whether an AI project is viable across technical, business, and operational dimensions before significant resources are committed. It helps identify potential risks and blockers early in the project lifecycle.

Explanation: Model training and hyperparameter tuning are part of the Model Development phase (Phase 5), not Data Preparation. Data preparation includes collection, cleaning, handling missing values, and feature engineering.

Explanation: Transfer learning leverages knowledge (typically pre-trained weights) from a model trained on a large dataset and applies it to a related but different problem. This can significantly reduce training time and data requirements.

Explanation: The F1 score is the harmonic mean of precision and recall. It provides a balanced measure when there is an uneven class distribution, combining both metrics into a single score.

Explanation: Model drift occurs when the statistical properties of the data or the relationship between features and target variables change over time, causing the model's predictions to become less accurate. This requires ongoing monitoring and potential retraining.

Explanation: The Value vs. Complexity Matrix helps organizations prioritize AI initiatives by plotting use cases based on their potential business value (y-axis) against their technical complexity (x-axis), enabling strategic portfolio decisions.

Explanation: Data augmentation applies transformations (rotations, flips, crops for images; synonyms, back-translation for text) to existing training samples, effectively increasing dataset size and diversity without collecting new data.

Explanation: Hyperparameters are settings that control the model training process itself (like learning rate, batch size, number of layers), not the learned weights. Tuning these systematically helps optimize model performance.