Thread by @Rami_Krispin

New release for @Uber forecasting library 🌈

Orbit is an open-source #Python library for Bayesian time series forecasting and inference applications developed by @UberEng 🧵 👇🏼

#timeseries #forecast #MachineLearning #PyTorch #Bayesian #Bayes

The library uses under the hood probabilistic programming languages with libraries such as #Python @mcmc_stan , Pyro, and #PyTorch to build the forecast estimators.

The new release, version 1.1 includes the following new features and changes: 👇🏼

New #forecasting model - Kernel Time-based #Regression (KTR). KTR model uses latent variables to define a smooth, time-varying representation of regression coefficients. Tutorials: 👇🏼
orbit-ml.readthedocs.io/en/latest/tuto…
orbit-ml.readthedocs.io/en/latest/tuto…
orbit-ml.readthedocs.io/en/latest/tuto…
orbit-ml.readthedocs.io/en/latest/tuto…

✨Improve the object-oriented workflow by changing the library's classes, defining three types:
- Forecaster
- Model Template
- Estimator
✨The change of the classes comes with some changes in the code syntax

✨Updating the model diagnostics tools adding supports for ArViz #Python library data visualization functions such as density, pair, and trace plots

See tutorial: orbit-ml.readthedocs.io/en/latest/tuto…

#dataviz #Datavisualization

Resources 👇🏼👇🏼👇🏼
Documentation: orbit-ml.readthedocs.io/en/latest/inde…
Source code: github.com/uber/orbit
Release blog: eng.uber.com/the-new-versio…
Colab notebook: colab.research.google.com/github/edwinng…

Supporting libraries
PyStan: github.com/stan-dev/pystan
Pyro: github.com/pyro-ppl/pyro
PyTorch: pytorch.org

LinkedIn post available here:
linkedin.com/feed/update/ur…