MLU-EXPLAIN

Visual explanations of core machine learning concepts


Machine Learning University (MLU) is an education initiative from Amazon designed to teach machine learning theory and practical application.

As part of that goal, MLU-Explain exists to teach important machine learning concepts through visual essays in a fun, informative, and accessible manner.

MLU Robot Deriving Beta Coefficient For Least Squares on Whiteboard

Explore Published Articles...



Neural Networks Article Image (A computational graph, with two input nodes, multiple nodes in some hidden layers, and two output nodes).

Neural Networks

Learn about neural networks, the backbone of many popular algorithms today, such as ChatGPT, Stable-Diffusion, and many others

Equality of Odds Article Image (A Scatterplot on the left with a decision boundary corresponding to a stacked column chart on the right).

Equality of Odds

Explore equality of odds, a metric used to quantify unfairness and remove bias from machine learning models.

Logistic Regression Article Image (A Scatterplot showing points for Sunny and Rainy days plotted by Temperature in degrees Fahrenheit and the predicted probability as a sigmoid curve).

Logistic Regression

Learn how logistic regression can be used for binary classification in machine learning through an interactive example.

Linear Regression Article Image (A Scatterplot showing orange points and a black line on the right. On the left math equations for The Normal Equation).

Linear Regression

Interactively learn about linear regression models as they're commonly used in the context of machine learning.

Reinforcement Learning Article Image (Grid with arrows, bananas, a robot, and a cactus).

Reinforcement Learning

Learn about Reinforcement Learning (RL) and the exploration-exploitation dilemma with this interactive article.

ROC & AUC Article Image (A Scatterplot showing three ROC curves: one labeled Perfect Classifier (line hugging left and top of plot), one labeled Our Classifier (bumpy line), and one labeled Random Classifier (diagonal line)).

ROC & AUC

A visual explanation of the Receiver Operating Characteristic Curve (ROC) curve, how it works with a live interactive example, and how it relates to Area Under The Curve (AUC).

Cross-Validation Article Image.

Cross-Validation

K-Fold Cross-Validation: a resampling technique to help improve estimates of test error rates compared to a simple validation set.

Train, Test, Validation Article Image (Groups of cats/dogs in circles).

Train, Test, And Validation Sets

Learn why it is best practice to split your data into training, testing, and validation sets, and explore the utility of each with a live machine learning model.

Precision Recall Article Image (Beeswarm Plot).

Precision & Recall

When it comes to evaluating classification models, accuracy is often a poor metric. This article covers two common alternatives, Precision and Recall, as well as the F1-score and Confusion Matrices.

Decision Tree Title Image

Random Forest

Learn how the majority vote and well-placed randomness can extend the decision tree model to one of machine learning's most widely-used algorithms, the Random Forest.

Decision Tree Title Image

Decision Trees

Explore one of machine learning's most popular supervised algorithms: the Decision Tree. Learn how the tree makes its splits, the concepts of Entropy and Information Gain, and why going too deep is problematic.

Bias Variance Title Image

The Bias Variance Tradeoff

Understand the tradeoff between under- and over-fitting models, how it relates to bias and variance, and explore interactive examples related to LOESS and KNN.

Double Descent Title Image

Double Descent: A Visual Introduction

Meet the double descent phenomenon in modern machine learning: what it is, how it relates to the bias-variance tradeoff, the importance of the interpolation regime, and a theory of what lies behind.

Double Descent 2 Title Image

Double Descent: A Mathematical Explanation

Deepen your understanding of the double descent phenomenon. The article builds on the cubic spline example introduced in Double Descent 1, describing in mathematical detail what is happening.