The answer
Some objectives can produce negative numerical values, depending on their definition. Common squared-error and standard cross-entropy losses are nonnegative under their usual assumptions.
How does a loss function measure the quality of a model’s prediction?
Can loss be negative?
More FAQs in How does a loss function measure the quality of a model’s prediction?
No. Accuracy counts correct decisions, while a loss can measure confidence and error size. Two models with equal accuracy can have different losses.
Validation loss measures fit on examples excluded from training updates. A growing gap between training and validation performance can indicate overfitting or distribution differences.
Still have questions?
Browse all our FAQs or reach out to our support team
Want to upskill yourself?
Your learning library
Small steps. Big possibilities.
Find your next course. Build something great.