Kesinlik
metric measuring how far the forecasts that the model says positive.
Definiteity (precision) is an metric that measures how much of the forecasts marked as "positive" of a precision classification model is true, and mathematically real positives are calculated by division of false positives with real positives. Simply express, certainty responds to the question "how much we can trust" when the model says positively. High precision means that the price of the model is low, that the majority of the positive examples are really positive.
The exactness is critical especially in scenarios where false positives are cost-effective or disturbing: for example, if a spam filter accidentally signs an email as spam, the user may miss an important email, so it is aimed at high accuracy in spam filters. However, it can lead to a large number of guesses that the model to optimize the exactness alone is treated very cautiously, therefore it can lead to drop of sensitivity; therefore, certainty is usually evaluated with compound metrics, such as precision-recall curve or F1 score. An informed balance can be installed between the pattern’s decision threshold and precision.
The exactness is critical especially in scenarios where false positives are cost-effective or disturbing: for example, if a spam filter accidentally signs an email as spam, the user may miss an important email, so it is aimed at high accuracy in spam filters. However, it can lead to a large number of guesses that the model to optimize the exactness alone is treated very cautiously, therefore it can lead to drop of sensitivity; therefore, certainty is usually evaluated with compound metrics, such as precision-recall curve or F1 score. An informed balance can be installed between the pattern’s decision threshold and precision.