AWS Certified Cloud Practitioner

Explanation

Accuracy is the correct evaluation metric for this scenario because:

1. Classification Problem: The task involves predicting plant diseases from photos, which is a classification problem (categorizing images into different disease classes).

2. Measuring Correct Predictions: Accuracy measures the proportion of correct predictions out of total predictions, which directly addresses the company's goal of evaluating "how many images the model classified correctly."

3. Why other options are incorrect:

   • R-squared score (A): Used for regression problems to measure how well the model explains variance in continuous data, not for classification.
   • Root mean squared error (RMSE) (C): Also used for regression problems to measure the average magnitude of prediction errors for continuous variables.
   • Learning rate (D): This is a hyperparameter used during model training to control how much to adjust model weights, not an evaluation metric.

4. Practical Application: For image classification tasks like this, accuracy is commonly used as a straightforward metric to understand overall model performance, especially when class distribution is balanced.

Note: While accuracy is appropriate here, for imbalanced datasets (where some diseases are rare), other metrics like precision, recall, or F1-score might provide better insights.