This is an implementation of Monotonic Calibrated Interpolated Look-Up Tables in TensorFlow.
These are fast-to-evaluate and interpretable lattice models, also known as interpolated look-up tables. This library also provides a rich and intuitive set of regularizations and monotonicity constraints configurable per feature.
It includes TensorFlow estimators for regression and classification with the most common set ups for lattice models:
- Calibrated Linear
- Calibrated Lattice
- Random Tiny Lattices (RTL)
- Embedded Tiny Lattices (ETL) (see Deep Lattice Networks and Partial Monotonic Functions)
Additionally this library provides two types of model components (or layers) that can be combined with other types of models (including neural networks):
- Calibration: piecewise linear calibration of signals.
- Lattice: interpolated look-up table implementation.
You can install our prebuilt pip package using
pip install tensorflow-latticebut please see the install section for more detailed instructions.
This tutorial contains more detailed explanation about lattice models and usage in TensorFlow, and check out API docs for python APIs.
TensorFlow Lattice is not an official Google product.