Скачать или смотреть Forecasting Hotel Demand and Cancellations with SARIMA Models | Time Series Analysis Project Demo

In this project demonstration, I present a live web application built to forecast hotel stays and cancellations using SARIMA modeling techniques. Developed as a solo project for my Time Series Analysis course, the app helps hotel teams optimize staffing, supply planning, and revenue strategies through precise time series forecasting.

The app was developed using Python (Flask, statsmodels, pandas, matplotlib) and deployed on PythonAnywhere for easy access. Key features include:

Dataset exploration and time series decomposition

Baseline and customized SARIMA forecasts

Model evaluation with RMSE, MAE, MAPE metrics

Residual diagnostics and model performance comparison

The goal is to provide an interactive platform for understanding, modeling, and improving hotel demand forecasts.

Live Demo covers:

Dataset exploration

Forecasting tool usage

Model performance evaluation

14-day property-level forecasts

Thanks for watching!