See data science before you code it
Neural networks, gradient descent, loss curves, decision trees — all animated in 3D before you write a single line of Python.
Animations included
6 concepts you'll see in 3D
Each animation runs in your browser. No video player — live WebGL rendered in real time.
Neural network forward pass
Watch weights activate layer by layer as data flows through a deep network in real time.
Gradient descent visualised
A 3D loss surface with a ball rolling to the minimum. Change the learning rate and watch the path change.
Decision tree splits
See how each split divides the feature space into regions. Visualise overfitting and pruning.
Confusion matrix live
A 2D animated heatmap that updates as the model trains. Precision, recall, F1 all shown spatially.
PCA dimensionality reduction
3D data cloud rotated and projected into 2D. Watch variance explained change in real time.
K-means clustering
Data points animated into clusters. Watch centroids move until convergence.
UEEP model — step 1
Understand before you build
The UEEP model starts with visual understanding. These animations are the Understand step — before any code, before any exercise, the concept must click visually.
Watch the Data Science animations
Open Animation Lab and select Data Science to start watching. Free, no sign-up needed.