☰
Free download

Table Of Contents

Built with 🛠 MkDocs - Theme 🖤 Github.

Classification

Explore Java classifiers that predict binary decisions and multiclass labels from structured data.

Decision boundary separating a naval mine from a rock Binary classification Sonar Rock or Mine Classification Classify reflected sonar measurements as rock or mine with a binary neural network. View walkthrough Email spam classification Binary classification Email Spam Detection Transform labeled email features into a probability and a final spam or not-spam decision. View walkthrough Credit-card fraud detection Binary classification Credit Card Fraud Detection Estimate whether an anonymized transaction is legitimate or fraudulent and inspect classification metrics. View walkthrough Decision regions separating three Iris flower classes Multiclass classification Iris Flower Classification Predict one of three Iris species from four measurements with a Softmax output layer. View walkthrough

Use classification when the target is a category, such as spam or not spam, fraud or legitimate, or one of several flower species.

Binary models usually use one Sigmoid output, while multiclass models use one Softmax probability for each class.