Sales Forecasting with Linear Regression

This project demonstrates a simple sales forecasting model using Linear Regression. It predicts total sales based on factors like price per unit, units sold, and operating profit.

Project Structure

.
├── README.md
├── app.py
├── data
│   ├── Adidas_US_Sales_Cleaned.csv
│   └── linear_regression_model.pkl
├── model.ipynb
├── requirements.txt
├── static
│   ├── app.js
│   └── style.css
└── templates
    └── index.html

Dataset

Data Source: https://www.kaggle.com/datasets/heemalichaudhari/adidas-sales-dataset

The dataset used in this project is Adidas_US_Sales_Cleaned.csv, which contains the following key columns:

• Price per Unit: Price of a single unit of a product.
• Units Sold: Number of units sold.
• Operating Profit: Profit generated from operations.
• Total Sales: The target variable for prediction, representing the total sales amount.

Model

The model used is a Linear Regression model, which aims to predict Total Sales based on the features Price per Unit, Units Sold, and Operating Profit.

Model Training and Testing

The data is split into training and testing sets using an 80-20 split. After training, predictions are made on the test set, and the performance is evaluated using Mean Squared Error (MSE).

Key Steps

1. Load the Dataset: The dataset is read from a CSV file located in the data folder.

   data = pd.read_csv('data/Adidas_US_Sales_Cleaned.csv')

2. Feature Selection: The relevant features for the model are selected:

   X = data[['Price per Unit', 'Units Sold', 'Operating Profit']]
   y = data['Total Sales']

3. Train-Test Split: The data is split into training and testing sets:

   X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

4. Model Training: The linear regression model is trained on the training data:

   model = LinearRegression()
   model.fit(X_train, y_train)

5. Model Prediction: Predictions are made on the test data:

   y_pred = model.predict(X_test)

6. Model Evaluation: The model's performance is evaluated using Mean Squared Error:

   mse = mean_squared_error(y_test, y_pred)

7. Model Saving: The trained model is saved as a .pkl file using joblib:

   joblib.dump(model, 'data/linear_regression_model.pkl')

Results

• Mean Squared Error (MSE): The MSE value represents the model's prediction error. The lower the MSE, the better the model's predictions.
• Model Coefficients: The coefficients of the linear regression model show the relationship between the features and the target variable.

Dependencies

Install the required Python packages using the following command:

pip install -r requirements.txt

How to Run the Code

1. Clone the repository:

   git clone https://github.com/Ismat-Samadov/Sales_Forecasting.git

2. Navigate to the project directory:

   cd Sales_Forecasting

3. Install the required dependencies:

   pip install -r requirements.txt

4. Run the Flask app:

   python app.py

5. Open a browser and go to http://localhost:5000 to use the web interface for sales prediction.