⚠️ Note

This project was originally developed in 2019. Some libraries, APIs, and functions used in the code may now be deprecated or updated.

It is recommended to use this repository primarily to understand the overall workflow, feature engineering techniques, and modeling approach.

For practical use, please update the dependencies and modify deprecated functions to align with the latest versions of the libraries.

🔄 Workflow

📌 Data Processing

• Load time series dataset
• Convert datetime and extract features (year, month, day, hour)
• Perform resampling (hourly, daily, weekly, monthly)
• Train-validation split

---

📊 Exploratory Analysis

• Trend and seasonality analysis
• Hypothesis-driven visualizations (year, month, hour, weekday/weekend)
• Time series decomposition
• Stationarity check using Augmented Dickey-Fuller test

---

🤖 Models Used

• Naive Forecasting
• Moving Average
• Simple Exponential Smoothing (SES)
• Holt’s Linear Trend Model
• Holt-Winters Model
• AR Model
• MA Model
• ARIMA
• SARIMA

Comparative analysis performed using RMSE (Root Mean Squared Error)

---

🛠️ Core Packages Used

• NumPy
• Pandas
• Matplotlib
• Statsmodels
• Scikit-learn

---