Share this article:
Source Towards Data Science

Darts’ Swiss Knife for Time Series Forecasting

Share This Article
Source Towards Data Science

A step-by-step tutorial on how to set up a single Python script with Darts — and then run your source time series through several alternative methods to find the best fit, tournament-like, among the models.

In this example, we will run five forecast methods on the source data and then compare their accuracy.

Darts embeds most of the widely used time series forecasting methods we know from more specialized Python packages. Among others, it wraps the facebook Prophet model and pmdarima’s auto-ARIMA package. Darts also implements several neural network architectures.

  • Exponential Smoothing
  • Naïve Drift + Seasonal Forecast
  • (auto-)SARIMAX
  • (V)ARIMA
  • Theta Method
  • facebook Prophet
  • Linear Regression
  • Recurrent Neural Networks such as vanilla RNN, LSTM, and GRU
  • Temporal Convolutional Network TCN
  • Fast Fourier Transform FFT
  • Neural Basis Expansion Analysis Time Series Forecasting N-BEATS
Article on Towards Data Science

Want to receive updates from us?

agree_checkbox *

By subscribing, you consent to Unit8 storing and processing the data provided above in order to provide you with the requested content. For more information, please review our Privacy Policy.

Our newsletter features industry news, the latest case studies, and future Unit8 events.


This page is only available in english