How to update the Prophet model for new data to capture the latest trends and seasonality? There are several sources that say that the Prophet model should be created newly for every new data. Is there any other efficient solution for model updation? As the model needs to handle huge amount of live data My aim is to automate the model for Anomaly detection of data. Please suggest if there are any other methodologies for automation of anomaly detection for live data.
I created a new model everytime for new data which took more time for the training as well as anomaly detection. I expect the model to take the recent trend of data into consideration. Sources state that the prophet model can be fit only once. Is there any other way for updating the model without training the entire data from the source for every new data that comes?
