Multiple views, one catalog

Browse by facet

Models do not belong to a single hierarchy. Change the lens to explore them by goal, data, family, mechanism, learning signal, behavior, or constraint.

What must the model produce?

classification

8 models

Assign one or more discrete labels to an input.

Decision TreeA Decision Tree predicts by following a sequence of feature-threshold questions to a leaf.Gradient BoostingGradient Boosting builds a sequence of small models, each correcting the current ensemble's mistakes.K-Nearest NeighborsKNN predicts from the labels of the most similar stored examples.Logistic RegressionLogistic Regression turns a weighted sum of features into a probability for a class.Multilayer PerceptronAn MLP stacks learned linear transformations and nonlinear activations to model complex feature interactions.Random ForestRandom Forest averages many deliberately varied decision trees to create a more stable predictor.Temporal Convolutional NetworkA TCN uses causal, usually dilated one-dimensional convolutions to model sequences with a large and controllable history.TransformerA Transformer builds context-aware representations by letting positions assign attention to other positions.

clustering

1 model

Discover groups of similar examples without predefined labels.

K-MeansK-Means divides points into $k$ groups by repeatedly assigning each point to its nearest center and moving centers to the assigned averages.

dimensionality reduction

1 model

Compress data into fewer variables while preserving useful structure.

Principal Component AnalysisPCA rotates data onto new perpendicular axes ordered by how much variance they capture.

forecasting

2 models

Predict future values from ordered historical observations.

Temporal Convolutional NetworkA TCN uses causal, usually dilated one-dimensional convolutions to model sequences with a large and controllable history.TransformerA Transformer builds context-aware representations by letting positions assign attention to other positions.

generation

2 models

Create new samples that resemble the data learned during training.

Diffusion ModelA diffusion model learns to turn noise into structured data by repeatedly reversing a gradual corruption process.TransformerA Transformer builds context-aware representations by letting positions assign attention to other positions.

regression

8 models

Predict a continuous numerical value.

Decision TreeA Decision Tree predicts by following a sequence of feature-threshold questions to a leaf.Gradient BoostingGradient Boosting builds a sequence of small models, each correcting the current ensemble's mistakes.K-Nearest NeighborsKNN predicts from the labels of the most similar stored examples.Linear RegressionLinear Regression predicts a number by adding weighted feature values and a baseline offset.Multilayer PerceptronAn MLP stacks learned linear transformations and nonlinear activations to model complex feature interactions.Random ForestRandom Forest averages many deliberately varied decision trees to create a more stable predictor.Temporal Convolutional NetworkA TCN uses causal, usually dilated one-dimensional convolutions to model sequences with a large and controllable history.TransformerA Transformer builds context-aware representations by letting positions assign attention to other positions.

representation

5 models

Learn useful features that make later tasks easier.

Diffusion ModelA diffusion model learns to turn noise into structured data by repeatedly reversing a gradual corruption process.Multilayer PerceptronAn MLP stacks learned linear transformations and nonlinear activations to model complex feature interactions.Principal Component AnalysisPCA rotates data onto new perpendicular axes ordered by how much variance they capture.Temporal Convolutional NetworkA TCN uses causal, usually dilated one-dimensional convolutions to model sequences with a large and controllable history.TransformerA Transformer builds context-aware representations by letting positions assign attention to other positions.