Tutorial
Create a confusion matrix with Python
Use scikit-learn to create a confusion matrix for a simple binary classification problem
On this page
A confusion matrix (or error matrix) is a visualization method for classifier algorithm results. More specifically, a confusion matrix is a table that breaks down the number of ground truth instances of a specific class against the number of predicted class instances. Confusion matrices are one of several evaluation metrics that are used to measure the performance of a classification model in machine learning. They can be used to calculate a number of other model performance metrics, such as precision and recall, among others.
Confusion matrices can be used with any classifier algorithm, such as Naïve Bayes, logistic regression models, and decision trees. Because of their wide applicability in data science and machine learning models, many packages and libraries come preinstalled with functions for creating confusion matrices, such as the sklearn.metrics module in scikit-learn for Python.
In this tutorial, you use scikit-learn to walk through creating a confusion matrix for a simple binary classification problem. The tutorial also covers how to compute classifier metrics, such as precision and F1 score.
Confusion matrix layout
In a confusion matrix, columns represent the predicted values of a given class while rows represent the actual values (that is, the ground truth) of a given class, or vice versa. The reverse also appears in research. This grid structure is a convenient tool for visualizing model classification accuracy by displaying the number of correct and incorrect predictions for all classes alongside one another.
A standard confusion matrix template for a binary classifier is divided into four quadrants:
The top-left box provides the number of true positives (TP), being the number of correct predictions for the positive class. The box beneath it is false positives (FP), those negative-class instances incorrectly identified as positive cases. These are called type I errors in statistics. The top-right box is the number of false negatives (FN), the actual positive instances erroneously predicted negative. Finally, the bottom-right box displays the number of true negatives (TN), which are actual negative class instances accurately predicted as negative. Totaling up each of these values provides the model’s total number of predictions.
Totaling up each of these values provides the model’s total number of predictions.
Prerequisites
You need an IBM Cloud account to create a watsonx.ai project.
Steps
Step 1. Set up your environment
While you can choose from several tools, this tutorial walks you through how to set up an IBM account to use a Jupyter Notebook. Jupyter Notebooks are widely used within data science to combine code, text, images, and data visualizations to formulate a well-formed analysis.
Log in to watsonx.ai using your IBM Cloud account.
Create a watsonx.ai project.
Create a Jupyter Notebook.
A notebook environment opens for you to load your data set and copy code from this tutorial to tackle a simple data visualization task. To view how each block of code works, each step’s code block is best inserted as a separate cell of code in your watsonx.ai project notebook.
Step 2. Install and import relevant libraries
You need a few libraries for this tutorial. Make sure to import the necessary Python libraries. If they're not installed, you can resolve this with a quick pip install (included at the beginning of the following code).
# install and import necessary libraries
!pip install ucimlrepo
!pip install sklearn
!pip install numpy
!pip install seaborn
import numpy as np
from sklearn import metrics
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score
from sklearn.metrics import classification_report
import matplotlib.pyplot as plt
import seaborn as sns
Step 3. Load the data
For this tutorial, you do not load data from an outside repository. The confusion matrix is a data visualization tool for evaluating classification algorithm performance. As such, your data is the actual and predicted values for a classification task. The actual values (or ground truth) are the manually created class labels for a data set. The predicted values are the predicted class labels for that same data set.
The ground truth and predicted values that are used for this tutorial are:
# load ground truth and predicted values from classier output and store as variables
ground_truth = [0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 1, 0, 0, 1]
predictions = [0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0]
These two lists are derived from a classifier model that is created with the Multinomial Naive Bayes tutorial. The Naïve Bayes tutorial uses a spam data set hosted by UCI Machine Learning Repository to provide step-by-step instructions on creating several iterations of a classifier.
This tutorial's predictions were obtained from the Naive Bayes tutorial's model by using stemming, no stopword removal, and TfidfVectorizer(). This is the same model predictions set that is displayed in the Naive Bayes tutorial's confusion matrix. You can compare how other models perform using confusion matrices by replacing the defined lists in this tutorial with other models' list-formatted output.
In the data for this tutorial, 0 signifies a data sample that is classified as "not spam" and 1 represents a data sample classified as "spam."
Step 4. Create confusion matrix
This step defines, then prints, a simple confusion matrix using the loaded list values. In the confusion_matrix() function, the first variable is the true label distribution and the second is the predicted label distribution.
# define confusion matrix
confusion_matrix = metrics.confusion_matrix(ground_truth, predictions)
print(confusion_matrix)
The code outputs a rudimentary confusion matrix:
Here, the upper-left value is the number of true positives, being the number of instances that are correctly classified as not spam in the test data. The upper-right is the number of false negatives, or those instances that are incorrectly classified as spam. The lower-left is the number of false positives that the spam data samples erroneously classified as not spam. The lower-right is the number of true negatives, which are those instances of spam that are correctly classified as spam.
Given that true positives and true negatives collectively outnumber false predictions in this confusion matrix, the classifier appears to perform well. However, as mentioned, this confusion matrix displays the actual number of correct and incorrect predictions for each class. What if you instead want to see each class's percentage of correct and incorrect predictions?
Step 5. Normalize confusion matrix
The scikit-learn confusion_matrix function lets you display the percentage of correct predictions in the test set rather than the number of actual and predicted class instances. To do so, you can insert the normalize parameter with the value all as the final term in the confusion_matrix function.
# define normalized confusion matrix
confusion_matrix_norm = metrics.confusion_matrix(ground_truth, predictions, normalize="all")
print(confusion_matrix_norm)
This returns the confusion matrix:
The normalize parameter can take three values: all, pred, and true. The first produces a confusion matrix in which each box displays the percentage for each classification type out of the total number of samples. For example, normalize="all" returns a confusion matrix with .8633 in the upper-left box. This means that approximately 86.33% of all the data instances that are classified by the model were correctly classified as instances of non-spam. All of the boxes in this confusion matrix should total 1.00 (or 100%).
The true parameter value normalizes the confusion matrix along the true class axis. This means that each row will total 100%. For your spam classification task, the top row shows what percentage of non-spam instances are correctly classified as non-spam (upper-left box) and which are incorrectly classified as spam (upper-right box). Similarly, the lower-right box shows what percentage of spam samples are correctly classified as such while the lower-left box displays the percentage of spam samples misclassified as non-spam.
The pred parameter value normalizes the confusion matrix along its predicted class axis. This normalization creates a confusion matrix in which each column totals 100%. The upper-left box is the proportion of predicted instances of non-spam that are correct, while the lower-left box is the percentage of predicted instances of non-spam that are incorrect. The right column correspondingly displays the proportion of predicted instances of spam that are correct and incorrect.
Step 6. Visualize confusion matrix
Python’s print command provides a basic confusion matrix for rudimentary model evaluation. But sometimes, you might want a more robust confusion matrix with labels. The scikit-learn metrics module contains the ConfusionMatrixDisplay function, which you can use to create a more detailed matrix and define labels. The following example shows the code for visualizing the confusion matrix using sklearn and the pyplot module.
# visualize confusion matrix with sklearn metrics
# define matrix and create labels
cm_display = metrics.ConfusionMatrixDisplay(confusion_matrix = confusion_matrix, display_labels = ["Not Spam", "Spam"])
# display matrix
cm_display.plot()
plt.show()
The following image shows the confusion matrix that it plots.
Another common method for visualizing a detailed confusion matrix is the seaborn library’s heatmap function. Note that when visualizing the matrix using heatmap, you also need to explicitly define the y- and x-axis labels with pyplot. Also, each column and row must be defined in the heatmap function. In this way, seaborn provides more control over the specifics of the final matrix output.
# define and print matrix with labels
sns.heatmap(confusion_matrix,
annot=True,
fmt='g',
xticklabels=['Not Spam','Spam'],
yticklabels=['Not Spam','Spam'])
# display matrix
plt.ylabel('Actual',fontsize=12)
plt.xlabel('Prediction',fontsize=12)
plt.show()
The following image shows the confusion matrix that it plots.
Note that you can also display the normalized confusion matrices by replacing confusion_matrix in these bits of code with confusion_matrix_norm.
Step 7. Computer evaluation metrics
The confusion matrix provides a readily accessible visualization of classifier predictive results. Thus, it is useful for calculating an array of classifier model evaluation metrics. You can use the sklearn.metrics module to compute some basic classifier evaluation measurements.
Accuracy
Model accuracy is perhaps the most basic evaluation metric. You can compute it using two previously defined lists with the following command.
# compute accuracy using sklearn.metrics
accuracy = accuracy_score(ground_truth, predictions)
print("Accuracy:", accuracy)
Be aware that accuracy is not the most informative metric. Imagine that you run a classifier on a data set of 100 instances. The model’s confusion matrix shows only one false negative and no false positives. For your spam filtering task, this accuracy of 99% is desirable. But, what if your model aims to classify highly contagious diseases? That 1% misclassification poses an enormous risk. Thus, you can use other evaluation metrics to provide a better picture of the performance of a classification algorithm.
Precision and recall
Precision, which is also called positive predicted value (PPV), signifies the proportion of predicted positive samples that actually belong to the class in question. Recall, which is also known as sensitivity or true positive rate (TPR), is the percentage of actual class instances that are correctly classified by the model. They are represented by the equations Precision = TP / TP+FP and Recall = TP / TP+FN.
You can compute them with the two different functions.
# compute precision using sklearn.metrics
precision = precision_score(ground_truth, predictions)
print("Precision:", precision)
# compute recall using sklearn.metrics
recall = recall_score(ground_truth, predictions)
print("Recall:", recall)
F1 score
Precision and recall can share an inverse relationship. As a model’s recall increases, returning more actual class instances (that is, true positives), the model inevitably decreases precision by misclassifying noninstances (that is, false positives). The F1 score attempts to counter this tradeoff by combining precision and recall, and the F1 score is particularly suitable for imbalanced data sets.
The F1 score is also called F-score, F-measure, or the harmonic mean of precision and recall. It represents a model’s total class-wise accuracy. The F1 score is represented by the equation F = 2PR / P+R, where P denotes precision and R denotes recall.
You can compute the F1 score for your spam filter predictions with the following command.
# compute F1 score using sklearn.metrics
F1 = f1_score(ground_truth, predictions)
print("F1 Score:", F1)
Classification report
You can compute all of these values simultaneously by using the scikit-learn classification_report function.
# create and print classification report
report = classification_report(ground_truth, predictions)
print(report)
Summary and next steps
This beginner tutorial provides an overview for how to use scikit-learn to create a confusion matrix and compute common evaluation metrics in Python for a simple binary classification problem. The functions that are used in this tutorial can also create confusion matrices for multiclass classification problems. Note that there are various other evaluation visualizations, such as ROC curve and AUC, for measuring the quality of classification models.
Try watsonx for free
Build an AI strategy for your business on one collaborative AI and data platform called IBM watsonx, which brings together new generative AI capabilities, powered by foundation models, and traditional machine learning into a powerful platform spanning the AI lifecycle. With watsonx.ai, you can train, validate, tune, and deploy models with ease and build AI applications in a fraction of the time with a fraction of the data.
Try watsonx.ai, the next-generation studio for AI builders.
Next steps
Explore more articles and tutorials about watsonx on IBM Developer.
Also, to learn more about classification or regression algorithms using Python and scikit-learn, see these tutorials in the Getting started with machine learning learning path: