Multiclass classification using scikit-learn

 

Multiclass classification is a type of supervised learning problem where the objective is to classify data into more than two distinct classes. This problem is ubiquitous in many fields, including image recognition, natural language processing, and sentiment analysis. In this article, we will explore how to perform multiclass classification using scikit-learn, a popular machine learning library in Python.

Data preparation

Before we can build a multiclass classification model, we need to prepare our data. Scikit-learn provides several datasets for practicing classification problems. For this article, we will use the iris dataset, which contains 150 samples of iris flowers, each with four features (sepal length, sepal width, petal length, and petal width) and a target variable indicating the species of the iris (setosa, versicolor, or virginica).

To load the iris dataset, we can use the following code:

from sklearn.datasets import load_iris

iris = load_iris()

X = iris.data

y = iris.target

Here, X contains the feature values, and y contains the target values. We can split the data into training and testing sets using the train_test_split function from scikit-learn. We can use 80% of the data for training and 20% for testing.

from sklearn.model_selection import train_test_split

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

Model training

We will use the Support Vector Machine (SVM) algorithm for our multiclass classification problem. SVM is a powerful algorithm that can work well for multiclass problems. We can use scikit-learn's SVC class to train an SVM model. We can create an instance of the SVC class with the desired hyperparameters and then fit the model to our training data.

from sklearn.svm import SVC

svm_model = SVC(kernel='linear', C=1, decision_function_shape='ovr')

svm_model.fit(X_train, y_train)

Here, we are using a linear kernel with a regularization parameter C of 1. We are also using the "one-vs-rest" (ovr) strategy for multiclass classification, which trains n_classes binary classifiers, one for each class.

Model evaluation

To evaluate our model's performance, we can use scikit-learn's accuracy_score function. This function computes the accuracy of the model on the test set.

from sklearn.metrics import accuracy_score

y_pred = svm_model.predict(X_test)

accuracy = accuracy_score(y_test, y_pred)

print("Accuracy:", accuracy)

Our model achieved an accuracy of 1.0, which means it correctly classified all the test instances.

Conclusion

In this article, we explored how to perform multiclass classification using scikit-learn. We used the iris dataset to demonstrate the process of preparing the data, training the model, and evaluating its performance. Scikit-learn provides many other classifiers and evaluation metrics for multiclass classification, and it is a powerful tool for building machine learning models.