Performance measures
Performance measures are used to evaluate learning algorithms and form an important aspect of machine learning. In some cases, these measures are also used as heuristics to build learning models.
Now let's explore the concept of the Probably Approximately Correct (PAC) theory. While we describe the accuracy of hypothesis, we usually talk about two types of uncertainties as per the PAC theory:
- Approximate: This measures the extent to which an error is accepted for a hypothesis
- Probability: This measure is the percentage certainty of the hypothesis being correct
The following graph shows how the number of samples grow with error, probability, and hypothesis:
Is the solution good?
The error measures for a classification and prediction ...
Get Practical Machine Learning now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
