Leveraging NeuralProphet and Nx for Time Series Forecasting in Elixir

In this presentation, George Guimarães explains how Elixir has evolved into a robust ecosystem for productive Machine Learning applications. He focuses on how to utilize NeuralProphet, a project developed at Stanford, to forecast time series data relevant to businesses, such as sales and user behavior. One of the key advantages of NeuralProphet, alongside its predecessor Prophet, is its capability to convey models in an understandable manner, with the use of simple terms and visualizations. Additionally, NeuralProphet operates using PyTorch, which allows the integration of the Nx ecosystem to enhance model performance and utility.