Articles → MACHINE LEARNING → Classification Report In Machine Learning

Classification Report In Machine Learning

This article describes the classification report in Machine Learning.

Purpose

A classification report is a performance evaluation metric that gives a detailed breakdown of how well a classification model is performing

Components Of A Classification Report

The classification report has the following components: -

Precision
Recall
F1-Score
Support

Let us discuss each component.

Precision

Precision specifies, out of all the positive predictions that the model has made, how many of them are correct?

The formula is: -

Parameter	Description
TP (True Positive)	The model correctly predicted positive.
FP (False Positive)	The model incorrectly predicted positive.

High precision is equivalent to a low number of false positives. For example, if your model predicts cancer for 100 patients, but only 80 have cancer. The precision is 80%.

Recall

Actual Positive
Actual Positive means the number of cases in your dataset that truly belong to the positive class, regardless of what the model predicts.

Recall measures how well a model finds all the actual positives in the dataset.

The formula is: -

Picture showing the formula for calculating recall

Parameter	Description
FN (False Negative)	Actual positives that the model missed

Consider a scenario where we build a model to detect whether a patient has cancer.

Actual positives (patients with cancer) = 100
Model Predicts:
      80 correctly detected (TP = 80)
      20 missed (FN = 20)

Picture showing calculation of recall using an example

So, the model caught 80% of cancer cases.

F1-Score

Support

It simply counts how many samples belong to each class, independent of predictions.

Posted By -	Karan Gupta

Posted On -	Friday, October 17, 2025

Query/Feedback

Your Email Id		**

Subject		*

Query/Feedback	Characters remaining 250	**