Tensorflow.js is an open-source library that is developed by Google for running machine learning models as well as deep learning neural networks in the browser or node environment.
The .confusionMatrix() function is used to calculate the confusion matrix from the stated true labels coupled with predicted labels.
Syntax:
tf.confusionMatrix(labels, predictions, numClasses)
Parameters:
- labels: It is the stated target labels that are supposed to be a zero based integers in favor of the classes. It has shape [numExamples]. Where, numExamples is the measure of incorporated instances. It can be of type tf.Tensor1D, TypedArray, or an array.
- predictions: It is the stated predicted classes that are supposed to be a zero based integers in favor of the classes. It should have shape equivalent to the stated labels. It can be of type tf.Tensor1D, TypedArray, or an array.
- numClasses: It is the number of total classes of type integer. Moreover, its measure should be greater than the greatest element in the stated labels as well as predictions. It is of type number.
Return Value: It returns tf.Tensor2D object.
Example 1:
Javascript
// Importing the tensorflow.js library import * as tf from "@tensorflow/tfjs" // Defining predictions, labels and // numClasses const lab = tf.tensor1d([3, 4, 1, 0, 1], 'int32' ); const pred = tf.tensor1d([1, 3, 0, 4, 1], 'int32' ); const num_Cls = 2; // Calling tf.confusionMatrix() method const output = tf.math.confusionMatrix(lab, pred, num_Cls); // Printing output output.print(); |
Output:
Tensor [[0, 0], [1, 1]]
Example 2:
Javascript
// Importing the tensorflow.js library import * as tf from "@tensorflow/tfjs" // Calling tf.confusionMatrix() method const res = tf.math.confusionMatrix( tf.tensor1d([3.3, 4.5, null , 'a' , 'b' ]), tf.tensor1d([-2, 5.3, -0.1, 4.3, 12.5]), 4 ); // Printing output res.print(); |
Output:
Tensor [[1, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0]]
Reference: https://js.tensorflow.org/api/latest/#confusionMatrix