Learn Before
Common Performance Metrics for Classification
Bayes error rate
In statistical classification, Bayes error rate is the lowest possible error rate for any classifier of a random outcome (into, for example, one of two categories) and is analogous to the irreducible error.
For deep learning, human-level performance can be regarded approximately as Bayes error rate.
0
1
Contributors are:
Who are from:
Tags
Data Science
Related
Confusion Matrix
ROC Curve and ROC AUC
Precision and Recall performance metrics.
F1 Score
Optimizing Criteria in Classification Problems
Satisficing Criteria in Classification Problems
Bayes error rate
What evaluation metric would you want to maximize based on the following scenario?
Recall of a Classification Model
Precision of a Classification Model
Sensitivity Analysis of a Classification Model
Learning Curve of a Classification Model
Having three evaluation metrics makes it harder for you to quickly choose between two different algorithms, and will slow down the speed with which your team can iterate. True/False?
If you had the four following models, which one would you choose based on the following accuracy, runtime, and memory size criteria?
Coverage
How to choose between precision and recall?
F-Measure
Learn After
Bayes Error Rate for (Naive) Bayes Classifier
Human Level Performance
Avoidable bias
If your goal is to have “human-level performance” be a proxy (or estimate) for Bayes error, how would you define “human-level performance”?
Human Level Proxy for Bayes Error
Which of the following statements about algorithmic performance do you agree with?
Bayes Calssifier and Bayes Error Rate