📈 Daily Active Users (DAU) Forecasting

An end-to-end time series analysis and forecasting project designed to predict Daily Active Users (DAU) using historical data. This project implements a SARIMA (Seasonal Autoregressive Integrated Moving Average) model to capture both trends and weekly seasonality in user engagement.

📋 Table of Contents

• Overview
• Project Structure
• Installation
• Usage
• Methodology & Results
• Technologies Used

---

📖 Overview

Predicting active users is crucial for understanding platform growth and managing resources. This repository contains the data preparation, exploratory data analysis, stationarity testing (ADF Test), and the training of a SARIMAX model to forecast DAU.

Key Features:

• Automated time series formatting and handling of missing dates.
• Custom Grid Search for hyperparameter tuning.
• Static and dynamic forecasting with confidence intervals.
• Performance evaluation using sMAPE, MAPE, MSE, and RMSE.

📂 Project Structure

├── data.csv # Raw historical DAU data ├── SARIMA.ipynb # Main Jupyter Notebook with analysis and modeling ├── Prophet.ipynb # Main Jupyter Notebook with analysis and modeling ├── requirements.txt # Python dependencies (optional) └── README.md # Project documentation