holiday effects

Model Perspective

Aug 23, 2022 · Fundamentals

How Prophet Implements Time Series Decomposition and Trend Modeling

This article explains Prophet’s algorithmic approach to time‑series forecasting, covering decomposition into trend, seasonality, holidays and error components, logistic and piecewise linear trend models, automatic change‑point detection, Fourier‑based seasonality, holiday handling, model fitting with PyStan, and practical Python code examples.

ProphetPythonholiday effects

0 likes · 12 min read

How Prophet Implements Time Series Decomposition and Trend Modeling

iQIYI Technical Product Team

Dec 20, 2019 · Artificial Intelligence

Advertising Inventory Forecasting Using an LSTM-Based Deep Learning Model

The iQIYI advertising team introduced an LSTM‑based deep‑learning model that forecasts inventory by normalizing data, clustering dimensions, and embedding fine‑grained holiday features, achieving significantly lower bias than their Adaptive‑ARIMA baseline and improving generalization while reducing training resources.

Advertising ForecastingLSTMdeep learning

0 likes · 10 min read

Advertising Inventory Forecasting Using an LSTM-Based Deep Learning Model