---

Time Series Forecasting with Python

Overview

This repository showcases a collection of time series forecasting projects implemented in Python. It covers a variety of models and techniques, including AR, ARIMA, SARIMA, VAR, Exponential Smoothing, Moving Average, Holt-Winters methods, and STL Decomposition. These projects demonstrate forecasting trends, seasonal patterns, and dependencies in time series data using real-world datasets.

---

Repository structure

This repository contains time series analysis using various models:

• Time Series Smoothing
• AR model (AutoRegressive)
• ARIMA (AutoRegressive Integrated Moving Average)
• SARIMA (Seasonal ARIMA)
• VAR (Vector AutoRegression)

---

How to Run the Project

This project is developed in Google Colab. Follow these steps to run the notebooks:

1. Clone or download the repository:

   git clone https://github.com/your-username/time-series-forecasting.git
   cd time-series-forecasting

2. Open Google Colab (https://colab.research.google.com/).

3. Upload a notebook file (e.g., AR_model.ipynb) by clicking on File > Upload Notebook.

4. Upload the corresponding dataset from the datasets/ folder into the Colab environment.

5. Update the dataset file path in the notebook if necessary.

6. Run the notebook cells sequentially to reproduce the results.

---

Models Included

1. AR Model (1_Daily_minimum_temps.csv)

• Description: AutoRegressive (AR) models forecast time series data based on past observations. It captures the dependency between a current observation and its lagged values.
• Dataset: Daily minimum temperature dataset.
• Objective: Predict future minimum temperatures using an AR model.

---

2. ARIMA Model (MaunaLoaDailyTemps.csv)

• Description: The ARIMA model combines autoregression (AR), differencing (I), and moving average (MA) to handle non-stationary data and trends.
• Dataset: Daily temperature readings from the Mauna Loa Observatory.
• Objective: Forecast future temperatures using ARIMA.

---

3. SARIMA Model (airline-passenger-traffic(1).csv)

• Description: Seasonal ARIMA (SARIMA) extends ARIMA by including seasonal components, ideal for data with clear seasonality.
• Dataset: Monthly airline passenger traffic data.
• Objective: Forecast seasonal passenger traffic.

---

4. VAR Model (PCEPersonalSpending.csv, M2SLMoneyStock.csv)

• Description: Vector AutoRegression (VAR) is a multivariate model that captures dependencies across multiple time series.
• Dataset: U.S. Personal Consumption Expenditure and Money Stock data.
• Objective: Analyze interdependencies and forecast future values of these economic indicators.

---

5. Exponential Smoothing (Electric_Production.csv)

• Description: Exponential smoothing assigns exponentially decreasing weights to past observations, making it effective for forecasting data with trends or seasonality.
• Dataset: Monthly electricity production data.
• Objective: Smooth and forecast electricity production trends.

---

6. Moving Average Method (Electric_Production.csv)

• Description: A simple method that uses the average of a fixed number of past observations to predict the next value.
• Dataset: Monthly electricity production data.
• Objective: Identify trends and forecast electricity production.

---

7. Holt-Winters and Holt’s Linear Methods (Electric_Production.csv)

• Description: Holt-Winters extends Holt’s Linear Trend model by adding a seasonal component, making it suitable for seasonal data.
• Dataset: Monthly electricity production data.
• Objective: Model seasonality and trends for improved forecasting.

---

8. STL Decomposition (Electric_Production.csv)

• Description: STL (Seasonal and Trend decomposition using Loess) separates a time series into seasonal, trend, and residual components.
• Dataset: Monthly electricity production data.
• Objective: Analyze and visualize the individual components of the time series.

---

Datasets

1. 1_Daily_minimum_temps.csv: Daily minimum temperature readings.
2. MaunaLoaDailyTemps.csv: Daily temperature readings from Mauna Loa.
3. airline-passenger-traffic(1).csv: Monthly airline passenger traffic data.
4. PCEPersonalSpending.csv: U.S. personal consumption expenditure data.
5. M2SLMoneyStock.csv: Money stock dataset.
6. Electric_Production.csv: Monthly electricity production data.

---

Results

Each notebook includes:

• Data preprocessing (e.g., handling missing values, stationarity checks).
• Model implementation and parameter tuning.
• Forecasts vs. actual values visualized through plots.
• Performance evaluation using metrics like RMSE and MAE.

---

Future Enhancements

• Add advanced models like LSTM and Facebook Prophet for forecasting.
• Develop an interactive dashboard for time series visualization and forecasting.

---

License

This project is licensed under the MIT License. See the LICENSE file for details.

---