Deep Neural Time Series
Forecasting a series with a deep network does not have to mean building an LSTM from scratch. This page follows one library, NeuralProphet, from its documentation to a quick guide and then to tuning it.
The same notes are in LTSM for time series.
The reference for NeuralProphet is NP Docs, the NeuralProphet class documentation. With the API known, Optimizing Time Series Forecasting with NeuralProphet: A Quick Guide is Miri Trope's quick guide to using it well. The last step is the settings: Hyper Param Tuning is Lavinia Guadagnolo on finding the optimal hyperparameters for a NeuralProphet forecast, which means knowing what each parameter does, where to set its values, and how they interact.
Last updated
Was this helpful?