ML VISUAL LAB / END-TO-END

Complete Machine Learning Workflow

All labs

Experiment workflow

5 stages from data to model

Sample dataset

Number of samples160

Data noise0.12

Number of features2

Manual data Enable click-to-add points on the canvas

New point class

The first two features are plotted. Extra features are still used by the model.

Configure data and preprocessing, then run training.

Configure the dataset and features. Turn on manual mode, then click the canvas to add points.

Understand This Experiment!

Machine learning is not just one model. A real workflow starts with data, preprocessing to make numbers comparable, training to find parameters, evaluation on unseen data, and finally a model ready to use.

Why split the data?

The model is tested on a test set that never took part in training. That is the most honest way to know whether a model learns patterns instead of memorizing data.

Connect with other labs

Explore decision boundaries deeper in Logistic Regression or edit Confusion Matrix directly.