A confusion matrix is a fundamental tool in the field of machine learning and data science, often used to assess the performance of classification models. It provides a detailed breakdown of the model’s predictions compared to the actual ground truth, allowing us to evaluate various aspects of model performance.
The confusion matrix has two specific parameters: actual and predicted values. It is primarily used in supervised learning, particularly for classification tasks, to assess the performance of machine learning models.
Confusion Matrix is a performance measurement for machine learning classification where output can be two or more classes. It is a table with 4 different combinations of predicted and actual values.
True Positive:
Interpretation: You predicted positive and it’s true.
True Negative:
Interpretation: You predicted negative and it’s true.
False Positive: (Type 1 Error)
Interpretation: You predicted positive and it’s false.
False Negative: (Type 2 Error)
Interpretation: You predicted negative and it’s false.
Recall – Out of all the positive classes, how many do we predict correctly? It should be as high as possible.
Recall = TP / TP+FN
Precision – Out of all the positive classes we have predicted correctly, how many are actually positive?
Precision = TP / TP+FP
Accuracy – Out of all the classes, how much we predicted correctly
F1-Score – It is difficult to compare two models with low precision and high recall or vice versa. So to make them comparable, we use F1-Score. F1-Score helps to measure Recall and Precision at the same time. It uses Harmonic Mean in place of Arithmetic Mean for an important reason, it provides a balance between precision and recall, particularly when dealing with imbalanced datasets. The harmonic mean penalizes extreme values more heavily, which is desirable when dealing with imbalanced datasets. It balances precision and recall, giving more importance to the lower of the two values
F1-Score = 2*Recall*Precision/ Recall + Precision
Specificity measures how exact the assignment to the positive class is.
Specificity = TN/FP+TN
Sensitivity measures how apt the model is to detect events in the positive class.
Sensitivity = TP/TP+FN
Type 1 Error (False Positive): Imagine you’re taking a medical test to check if you have a particular disease. A Type 1 error is like when the test mistakenly says you have the disease when you actually don’t. In other words, it’s a false alarm. It’s like a smoke detector in your kitchen going off when there’s no fire. It can lead to unnecessary worry or actions, like starting treatment when you’re not sick.
Type 2 Error (False Negative): Now, think about the same medical test. A Type 2 error is when the test misses that you actually have the disease, telling you that you’re healthy when you’re not. It’s like the smoke detector not going off when there’s a real fire in your kitchen. This can be even more dangerous because you might not get the treatment you need, thinking you’re fine when you’re not.
In summary, a Type 1 error is a false alarm, and a Type 2 error is a missed alarm. Both types of errors are important to consider when making decisions based on tests or data because they can have real consequences, especially in critical situations like medical diagnoses or security systems.