DY0-001: CompTIA DataAIDemo

Question

A data scientist has developed a model that estimates the likelihood of an error occurring in a factory. The model’s historical accuracy is 90%. For a particular factory, the model reports a likelihood score of 0.90. Which of the following describes a confidence score of 0.90?

A Running this model for all known factory issues, it is expected the model will identify 90 out of 100 known factory issues.
B Running this model on 100 samples of factories, a certain model performance is expected for 90 out of the 100 samples.
C Running this model 100 times on a factory, it is expected the model will predict 90 out of 100 factory errors.
D Running this model 100 times within a factory, it is expected the model will predict error 90 out of 100 times the model is ran.

Explanation

For binary classification, a prediction score represents the model’s certainty that a specific observation belongs to the positive class. A score of 0.90 for this factory therefore expresses a predicted error likelihood of 90%, which corresponds to predicting an error in approximately 90 of 100 equivalent prediction opportunities. Historical accuracy measures aggregate correctness across a dataset and is distinct from the score for one prediction.

Learn more

Binary Classification - Amazon Machine Learning

DY0-001: CompTIA DataAIDemo

AstroTutor

Notes

Readiness

QuestionQ61

Community Discussion