keepseatAI HANDBOOK
One catalog, two views

Browse models

Group models by the question you are asking, or search the complete catalog directly.

What must the model produce?

anomaly detection

1 model

Identify observations that differ unusually from the expected pattern.

ModernTCNModernTCN is a pure convolutional architecture that uses patches, large depthwise kernels, and separate temporal and cross-variable mixing for general time-series analysis.

classification

10 models

Assign one or more discrete labels to an input.

Convolutional Neural Network (CNN)A CNN stacks convolutional layers to turn pixels into features and, ultimately, a prediction.Decision TreeGradient BoostingK-Nearest NeighborsLogistic RegressionMultilayer PerceptronModernTCNModernTCN is a pure convolutional architecture that uses patches, large depthwise kernels, and separate temporal and cross-variable mixing for general time-series analysis.Random ForestTemporal Convolutional NetworkTransformer

clustering

1 model

Discover groups of similar examples without predefined labels.

K-Means

dimensionality reduction

1 model

Compress data into fewer variables while preserving useful structure.

Principal Component Analysis

forecasting

4 models

Predict future values from ordered historical observations.

Chronos ForecastingChronos is a family of pretrained time-series models that transfers patterns learned across many datasets into probabilistic forecasts for new series.ModernTCNModernTCN is a pure convolutional architecture that uses patches, large depthwise kernels, and separate temporal and cross-variable mixing for general time-series analysis.Temporal Convolutional NetworkTransformer

generation

2 models

Create new samples that resemble the data learned during training.

Diffusion ModelTransformer

regression

8 models

Predict a continuous numerical value.

Decision TreeGradient BoostingK-Nearest NeighborsLinear RegressionMultilayer PerceptronRandom ForestTemporal Convolutional NetworkTransformer

representation

7 models

Learn useful features that make later tasks easier.

Convolutional Neural Network (CNN)A CNN stacks convolutional layers to turn pixels into features and, ultimately, a prediction.Diffusion ModelMultilayer PerceptronModernTCNModernTCN is a pure convolutional architecture that uses patches, large depthwise kernels, and separate temporal and cross-variable mixing for general time-series analysis.Principal Component AnalysisTemporal Convolutional NetworkTransformer