Assign products to predefined categories stored as IDs 101, 202, and 303. These numbers identify categories; their order or differences do not represent quantities.
Separately, estimate a product price as an amount and round it to whole yen for display. Also draft a product description. Compare the meaning of these outputs.
Choose TWO appropriate descriptions of the tasks.
Distinguish tasks by output meaning rather than storage format: category IDs represent classification, price amounts can be regression targets, and description drafts are generated content. Evaluate each output separately.
Detailed explanation
These IDs have no quantitative ordering or distance meaning. Do not interpret their numeric distance directly as the severity of category errors.
Category correctness and draft factual accuracy are separate. Review generated content against verified product specifications using suitable evaluation.
The prediction target matters more than display format. Predicting predefined price bands would be classification; this scenario estimates the amount itself.
The target is a price quantity. Rounding for display does not turn it into predicting a predefined price-band class.
The numbers encode categories. Numeric storage alone does not turn the task into regression over a quantity.
Check official documentation
Compare the following points with the official documentation.
Review SageMaker AI classification and regression definitions and Bedrock classification and text-generation examples. Compare category IDs, price estimates, price-band classes, and description drafts by output meaning and evaluation method.
Official references
What to confirm
Distinguish categories, quantities, and generated text without deciding from numeric IDs or rounding alone. Do not use classification accuracy as factual verification of generated descriptions.
Key points
- Output meaning
- Category IDs
- Amounts and display
- Content evaluation
Notes
- Environment: Current AIF-C01 Task 1.1 and related AWS primary-source review using invented examples; no model training or inference is performed.
- Review SageMaker AI classification and regression definitions and Bedrock classification and text-generation examples. Compare category IDs, price estimates, price-band classes, and description drafts by output meaning and evaluation method.
- Distinguish categories, quantities, and generated text without deciding from numeric IDs or rounding alone. Do not use classification accuracy as factual verification of generated descriptions.
Foundation review
Separate codes from quantities
Category IDs 101, 202, and 303 are not quantities like price or temperature. A numeric representation does not make meaningless ID differences suitable prediction errors.
Separate targets from display formats
Rounding a price estimate differs from predicting a predefined price-band class. Decide between these tasks from the required output meaning, not solely from whether the display is an integer.
Evaluate generated content separately
Drafting descriptions and predicting categories or amounts require different outputs. Generative models can also perform classification; avoid treating capabilities as mutually exclusive and evaluate each output with appropriate ground truth or review.